技术领域Technical Field
本发明涉及计算机技术领域，尤其涉及一种驾驶员风险评价管理方法及系统。The present invention relates to the field of computer technology, and in particular to a driver risk assessment management method and system.
背景技术Background technique
驾驶员风险评价管理是一项关键的安全管理实践，其旨在通过有效的监测、评估和干预驾驶员行为，以此来降低交通事故发生的风险，提高道路安全性，在现有的驾驶员风险评价管理方法中，通常是较为笼统地将所有危险驾驶行为归为一个大类，而后便以该大类计算驾驶得分和评定等级，并未将危险驾驶行为进行具体的分类，并且对于驾驶得分和评定等级也具有较强的主观成分，导致其准确度不高，从而无法准确地选取对应的管控策略，同时在对于驾驶员的风险评价中，大多时候会忽略驾驶员的心理状况的评定得分，驾驶员的心理状况评定同样是影响其驾驶行为的重要因素，加入心理状况得分能更好地评定驾驶员的风险等级，从而能够更好地选取对应的管控策略对驾驶员进行管理。Driver risk assessment management is a key safety management practice, which aims to reduce the risk of traffic accidents and improve road safety by effectively monitoring, evaluating and intervening in driver behavior. In the existing driver risk assessment management methods, all dangerous driving behaviors are usually classified into one category in a general way, and then the driving scores and ratings are calculated based on this category. Dangerous driving behaviors are not specifically classified, and the driving scores and ratings also have a strong subjective component, resulting in low accuracy, making it impossible to accurately select the corresponding control strategy. At the same time, in the risk assessment of drivers, the assessment score of the driver's psychological condition is often ignored. The assessment of the driver's psychological condition is also an important factor affecting his driving behavior. Adding the psychological condition score can better assess the driver's risk level, so that the corresponding control strategy can be better selected to manage the driver.
发明内容Summary of the invention
本发明的目的在于克服现有技术的不足，本发明提供了一种驾驶员风险评价管理方法及系统，不仅能够更为准确且客观地评价驾驶员的综合风险得分，并且能够更好地选取对应的管控策略对驾驶员进行管理。The purpose of the present invention is to overcome the deficiencies of the prior art. The present invention provides a driver risk assessment management method and system, which can not only more accurately and objectively evaluate the driver's comprehensive risk score, but also better select corresponding control strategies to manage the driver.
为了解决上述技术问题，本发明提供了一种驾驶员风险评价管理方法，所述方法包括：In order to solve the above technical problems, the present invention provides a driver risk assessment management method, the method comprising:
获取驾驶员的驾驶行为数据，并对驾驶员的驾驶行为数据进行数据预处理，获得预处理后的驾驶行为数据；Acquiring the driving behavior data of the driver, and performing data preprocessing on the driving behavior data of the driver to obtain the preprocessed driving behavior data;
对预处理后的驾驶行为数据进行特征分析，并基于特征分析结果利用预设特征阈值标定各类危险驾驶行为；Perform feature analysis on the pre-processed driving behavior data, and use preset feature thresholds to calibrate various types of dangerous driving behaviors based on the feature analysis results;
基于各类危险驾驶行为的标定结果利用聚类分析法进行聚类，获得聚类结果；Cluster analysis is used to cluster the calibration results of various dangerous driving behaviors to obtain cluster results;
基于聚类结果计算类别权重，基于所述类别权重量化各类危险驾驶行为的相对风险，并基于所述类别权重和相对风险计算驾驶行为得分；Calculating category weights based on the clustering results, quantifying relative risks of various types of dangerous driving behaviors based on the category weights, and calculating driving behavior scores based on the category weights and relative risks;
获取驾驶员的心理素质数据，基于所述心理素质数据利用预设要素测评值计算驾驶员的心理状况得分，并基于所述心理状况得分和驾驶行为得分计算驾驶员的综合风险得分；Acquiring the driver's psychological quality data, calculating the driver's psychological condition score based on the psychological quality data using preset factor evaluation values, and calculating the driver's comprehensive risk score based on the psychological condition score and the driving behavior score;
基于驾驶员的综合风险得分对驾驶员进行风险等级评级，基于驾驶员的风险等级生成对应的管控策略，并将管控策略发送至对应的驾驶员。The driver's risk level is rated based on the driver's comprehensive risk score, a corresponding management and control strategy is generated based on the driver's risk level, and the management and control strategy is sent to the corresponding driver.
可选的，所述获取驾驶员的驾驶行为数据，并对驾驶员的驾驶行为数据进行数据预处理，获得预处理后的驾驶行为数据，包括：Optionally, the acquiring the driving behavior data of the driver and performing data preprocessing on the driving behavior data of the driver to obtain the preprocessed driving behavior data includes:
基于数据传感器采集驾驶员的驾驶行为数据；Collecting the driver's driving behavior data based on data sensors;
对所述驾驶行为数据进行数据异常筛除处理，获得数据异常筛除处理后的驾驶行为数据；Performing data anomaly screening processing on the driving behavior data to obtain driving behavior data after the data anomaly screening processing;
对数据异常筛除处理后的驾驶行为数据进行去冗处理，获得去冗处理后的驾驶行为数据；Performing redundancy processing on the driving behavior data after the data anomaly screening process to obtain the driving behavior data after the redundancy processing;
对去冗处理后的驾驶行为数据进行数据转换处理，获得预处理后的驾驶行为数据。The driving behavior data after redundancy removal is subjected to data conversion processing to obtain pre-processed driving behavior data.
可选的，所述对预处理后的驾驶行为数据进行特征分析，并基于特征分析结果利用预设特征阈值标定各类危险驾驶行为，包括：Optionally, the pre-processed driving behavior data is subjected to feature analysis, and various types of dangerous driving behaviors are calibrated using preset feature thresholds based on the feature analysis results, including:
对预处理后的驾驶行为数据进行分段，获得若干段驾驶行为分段数据；Segmenting the preprocessed driving behavior data to obtain a number of driving behavior segmented data;
对每一段驾驶行为分段数据计算对应的均值、最大值、最小值和标准差，并基于每一段驾驶行为分段数据对应的均值、最大值、最小值和标准差获得特征指标，所述特征指标包括最大行驶车速、最大行驶加速度、最大航向角速度、注视点位置均值、眨眼频率和瞳孔面积变化率；Calculating the corresponding mean, maximum, minimum and standard deviation for each segment of driving behavior data, and obtaining characteristic indicators based on the corresponding mean, maximum, minimum and standard deviation for each segment of driving behavior data, wherein the characteristic indicators include maximum driving speed, maximum driving acceleration, maximum heading angular velocity, mean of gaze point position, blinking frequency and pupil area change rate;
将所述特征指标与预设特征阈值进行比较，基于比较结果标定各类危险驾驶行为。The characteristic index is compared with a preset characteristic threshold, and various types of dangerous driving behaviors are calibrated based on the comparison result.
可选的，所述将所述特征指标与预设特征阈值进行比较，基于比较结果标定各类危险驾驶行为，包括Optionally, the characteristic index is compared with a preset characteristic threshold, and various types of dangerous driving behaviors are calibrated based on the comparison result, including
若所述最大行驶车速大于预设特征阈值中的最大车速阈值，则标定为超速驾驶；If the maximum driving speed is greater than the maximum speed threshold in the preset characteristic threshold, it is marked as speeding;
若所述最大行驶加速度大于预设特征阈值中的最大加速度阈值，则标定为急加速驾驶；If the maximum driving acceleration is greater than the maximum acceleration threshold in the preset characteristic threshold, it is marked as rapid acceleration driving;
若所述最大航向角速度大于预设特征阈值中的最大航向角速度阈值，则标定为急变道驾驶；If the maximum heading angular velocity is greater than the maximum heading angular velocity threshold in the preset characteristic threshold, it is marked as a sudden lane change driving;
若所述注视点位置均值大于预设特征阈值中的最大注视点偏移距离或所述注视点位置均值小于预设特征阈值中的最小注视点偏移距离，则标定为分心驾驶；If the mean of the gaze point positions is greater than the maximum gaze point offset distance in the preset feature threshold or the mean of the gaze point positions is less than the minimum gaze point offset distance in the preset feature threshold, it is marked as distracted driving;
若所述眨眼频率小于预设特征阈值中的最小眨眼频率和/或所述瞳孔面积变化率小于预设特征阈值中的最小瞳孔面积变化率，则标定为疲劳驾驶。If the blinking frequency is less than the minimum blinking frequency in the preset characteristic threshold and/or the pupil area change rate is less than the minimum pupil area change rate in the preset characteristic threshold, it is calibrated as fatigue driving.
可选的，所述基于各类危险驾驶行为的标定结果利用聚类分析法进行聚类，获得聚类结果，包括：Optionally, the calibration results based on various types of dangerous driving behaviors are clustered using a cluster analysis method to obtain cluster results, including:
计算每类危险驾驶行为的标定结果超出预设特征阈值的幅值，利用聚类分析法对所有超出预设特征阈值的幅值进行聚类，获得各类危险驾驶行为的分类占比。The amplitude of the calibration results of each type of dangerous driving behavior exceeding the preset feature threshold is calculated, and the cluster analysis method is used to cluster all the amplitudes exceeding the preset feature threshold to obtain the classification proportion of each type of dangerous driving behavior.
可选的，所述基于聚类结果计算类别权重，基于所述类别权重量化各类危险驾驶行为的相对风险，并基于所述类别权重和相对风险计算驾驶行为得分，包括：Optionally, calculating the category weights based on the clustering results, quantifying the relative risks of various types of dangerous driving behaviors based on the category weights, and calculating the driving behavior scores based on the category weights and the relative risks include:
基于聚类结果计算各类危险驾驶行为对应的初始评价指标；Calculate the initial evaluation index corresponding to each type of dangerous driving behavior based on the clustering results;
对对应的初始评价指标中的极小型指标进行正向化处理，获得评价正向化指标，并对所述评价正向化指标进行标准化处理，获得标准化正向矩阵；Performing positive processing on the extremely small indicators in the corresponding initial evaluation indicators to obtain positive evaluation indicators, and performing normalization processing on the positive evaluation indicators to obtain a normalized positive matrix;
基于所述标准化正向矩阵利用熵权法计算对应的类别权重；Calculate the corresponding category weights using the entropy weight method based on the standardized forward matrix;
基于对应的类别权重建立在预设单位时间中各类危险驾驶行为的单位距离风险，基于所述单位距离风险量化各类危险驾驶行为的相对风险；Establishing unit distance risks of various types of dangerous driving behaviors in a preset unit time based on corresponding category weights, and quantifying relative risks of various types of dangerous driving behaviors based on the unit distance risks;
对各类危险驾驶行为对应的类别权重和相对风险进行变量替换，基于变量替换后的类别权重和相对风险计算驾驶行为得分。The category weights and relative risks corresponding to various types of dangerous driving behaviors are replaced with variables, and the driving behavior scores are calculated based on the category weights and relative risks after variable replacement.
可选的，所述基于所述单位距离风险量化各类危险驾驶行为的相对风险，包括：Optionally, the quantifying the relative risks of various types of dangerous driving behaviors based on the risk per unit distance includes:
基于所述单位距离风险利用交叉指数法构建交叉指数矩阵；Based on the unit distance risk, a cross-index matrix is constructed using a cross-index method;
基于所述交叉指数矩阵量化各类危险驾驶行为的相对风险。The relative risks of various types of dangerous driving behaviors are quantified based on the cross-index matrix.
可选的，所述获取驾驶员的心理素质数据，基于所述心理素质数据利用预设要素测评值计算驾驶员的心理状况得分，并基于所述心理状况得分和驾驶行为得分计算驾驶员的综合风险得分，包括：Optionally, the acquiring of the driver's psychological quality data, calculating the driver's psychological condition score based on the psychological quality data using preset factor evaluation values, and calculating the driver's comprehensive risk score based on the psychological condition score and the driving behavior score, includes:
采集驾驶员的心理素质数据，所述心理素质数据包括心理认知能力、人格特质和心理健康状况；Collecting the driver's psychological quality data, including psychological cognitive ability, personality traits and mental health status;
对驾驶员的心理认知能力、人格特质和心理健康状况分别赋予对应的加权权重，基于对应的加权权重利用预设要素测评值计算驾驶员的心理状况得分；Assign corresponding weights to the driver's psychological cognitive ability, personality traits and mental health status respectively, and calculate the driver's psychological status score based on the corresponding weights using the preset factor evaluation values;
基于所述心理状况得分和驾驶行为得分利用预设得分比例计算驾驶员的综合风险得分。The driver's comprehensive risk score is calculated based on the psychological condition score and the driving behavior score using a preset score ratio.
可选的，所述基于驾驶员的综合风险得分对驾驶员进行风险等级评级，基于驾驶员的风险等级生成对应的管控策略，并将管控策略发送至对应的驾驶员，包括：Optionally, the risk level rating of the driver based on the driver's comprehensive risk score, generating a corresponding control strategy based on the driver's risk level, and sending the control strategy to the corresponding driver includes:
基于驾驶员的综合风险得分利用预设评级标准对驾驶员划分风险等级；Classify drivers into risk levels based on their comprehensive risk scores using preset rating criteria;
基于驾驶员的风险等级利用对驾驶员所标定的各类危险驾驶行为生成对应的管控策略，并将对应的管控策略实时发送至对应的驾驶员中。Based on the driver's risk level, corresponding control strategies are generated using the various dangerous driving behaviors calibrated for the driver, and the corresponding control strategies are sent to the corresponding drivers in real time.
另外，本发明还提供了一种驾驶员风险评价管理系统，所述系统包括：In addition, the present invention also provides a driver risk assessment management system, the system comprising:
数据预处理模块，用于获取驾驶员的驾驶行为数据，并对驾驶员的驾驶行为数据进行数据预处理，获得预处理后的驾驶行为数据；A data preprocessing module, used to obtain the driver's driving behavior data, and perform data preprocessing on the driver's driving behavior data to obtain the preprocessed driving behavior data;
危险驾驶行为标定模块，用于对预处理后的驾驶行为数据进行特征分析，并基于特征分析结果利用预设特征阈值标定各类危险驾驶行为；The dangerous driving behavior calibration module is used to perform feature analysis on the pre-processed driving behavior data and calibrate various types of dangerous driving behaviors based on the feature analysis results using preset feature thresholds;
聚类模块，用于基于各类危险驾驶行为的标定结果利用聚类分析法进行聚类，获得聚类结果；A clustering module is used to cluster the calibration results of various dangerous driving behaviors using a clustering analysis method to obtain clustering results;
驾驶行为得分计算模块，用于基于聚类结果计算类别权重，基于所述类别权重量化各类危险驾驶行为的相对风险，并基于所述类别权重和相对风险计算驾驶行为得分；A driving behavior score calculation module, used to calculate category weights based on clustering results, quantify relative risks of various types of dangerous driving behaviors based on the category weights, and calculate driving behavior scores based on the category weights and relative risks;
综合风险得分计算模块，用于获取驾驶员的心理素质数据，基于所述心理素质数据利用预设要素测评值计算驾驶员的心理状况得分，并基于所述心理状况得分和驾驶行为得分计算驾驶员的综合风险得分；A comprehensive risk score calculation module, used to obtain the driver's psychological quality data, calculate the driver's psychological condition score based on the psychological quality data using the preset factor evaluation value, and calculate the driver's comprehensive risk score based on the psychological condition score and the driving behavior score;
管控模块，用于基于驾驶员的综合风险得分对驾驶员进行风险等级评级，基于驾驶员的风险等级生成对应的管控策略，并将管控策略发送至对应的驾驶员。The control module is used to rate the risk level of the driver based on the driver's comprehensive risk score, generate a corresponding control strategy based on the driver's risk level, and send the control strategy to the corresponding driver.
在本发明实施例中，根据对驾驶行为数据的特征分析结果利用预设特征阈值标定各类危险行为，将危险驾驶行为进行具体的标定分类，再通过计算类别权重和危险驾驶行为的相对风险来计算驾驶行为得分，能够更为客观地评定计算驾驶员的驾驶行为得分，并将驾驶员的心理状况得分也加入至风险得分的考虑中，通过由心理状况得分和驾驶行为得分所生成的综合风险得分能够更为准确且客观地对驾驶员进行风险评价，从而能够更为准确且具体地选取对应的管控策略，对驾驶员的管理能够达到更好的效果。In an embodiment of the present invention, various types of dangerous behaviors are calibrated using preset characteristic thresholds based on the characteristic analysis results of driving behavior data, and the dangerous driving behaviors are specifically calibrated and classified. The driving behavior score is then calculated by calculating the category weight and the relative risk of the dangerous driving behavior. The driving behavior score of the driver can be evaluated and calculated more objectively, and the driver's psychological condition score is also added to the risk score consideration. The comprehensive risk score generated by the psychological condition score and the driving behavior score can more accurately and objectively evaluate the risk of the driver, so that the corresponding control strategy can be selected more accurately and specifically, and the management of the driver can achieve better results.
附图说明BRIEF DESCRIPTION OF THE DRAWINGS
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见的，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其它的附图。In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
图1是本发明实施例中的驾驶员风险评价管理方法的流程示意图；FIG1 is a flow chart of a driver risk assessment management method in an embodiment of the present invention;
图2是本发明实施例中的驾驶员风险评价管理系统的结构组成示意图。FIG. 2 is a schematic diagram of the structural composition of a driver risk assessment management system in an embodiment of the present invention.
具体实施方式Detailed ways
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其它实施例，都属于本发明保护的范围。The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
实施例一Embodiment 1
请参阅图1，图1是本发明实施例中的驾驶员风险评价管理方法的流程示意图。Please refer to FIG. 1 , which is a flow chart of a driver risk assessment management method in an embodiment of the present invention.
如图1所示，一种驾驶员风险评价管理方法，所述方法包括：As shown in FIG1 , a driver risk assessment management method includes:
S11：获取驾驶员的驾驶行为数据，并对驾驶员的驾驶行为数据进行数据预处理，获得预处理后的驾驶行为数据；S11: Acquire the driving behavior data of the driver, and perform data preprocessing on the driving behavior data of the driver to obtain the preprocessed driving behavior data;
在本发明具体实施过程中，所述获取驾驶员的驾驶行为数据，并对驾驶员的驾驶行为数据进行数据预处理，获得预处理后的驾驶行为数据，包括：基于数据传感器采集驾驶员的驾驶行为数据；对所述驾驶行为数据进行数据异常筛除处理，获得数据异常筛除处理后的驾驶行为数据；对数据异常筛除处理后的驾驶行为数据进行去冗处理，获得去冗处理后的驾驶行为数据；对去冗处理后的驾驶行为数据进行数据转换处理，获得预处理后的驾驶行为数据。In the specific implementation process of the present invention, the method of acquiring the driving behavior data of the driver and performing data preprocessing on the driving behavior data of the driver to obtain the preprocessed driving behavior data includes: collecting the driving behavior data of the driver based on a data sensor; performing data anomaly screening processing on the driving behavior data to obtain the driving behavior data after the data anomaly screening processing; performing redundancy processing on the driving behavior data after the data anomaly screening processing to obtain the driving behavior data after the redundancy processing; performing data conversion processing on the driving behavior data after the redundancy processing to obtain the preprocessed driving behavior data.
具体的，通过数据传感器采集驾驶员的驾驶行为数据，驾驶行为数据包括车辆运行数据和驾驶员的动作数据；对驾驶行为数据进行数据异常筛除处理，识别或删除离群点并解决不一致性来清理数据，清除异常数据，获得数据异常筛除处理后的驾驶行为数据；对数据异常筛除处理后的驾驶行为数据进行去冗处理，检查驾驶行为数据中是否存在重复的数据，若有重复数据，则删除其中的一条或多条数据，删除重复数据后便对驾驶行为数据进行缺失值的处理，对于数据缺失值的处理，识别数据中存在的缺失值并分析缺失值的类型和分布情况，根据缺失值的类型和分布情况采用适当的方法，若缺失值较少，则采用插值方法进行填补缺失值，若存在大量缺失值，则应采用平均数填补法对缺失值进行填补，直至将所有缺失值填补完成后便完成数据去冗处理，获得去冗处理后的驾驶行为数据；对去冗处理后的驾驶行为数据进行数据转换处理，将去冗处理后的驾驶行为数据转换成可读取的格式，并将其单位和时间格式进行统一化处理，完成统一化处理后，便对驾驶行为数据进行数据规范化处理，采用极大值-极小值规范化方法，将驾驶行为数据变化到新的特定区域空间中，按新特定区域空间中的数值将驾驶行为数据集转换数值类型，完成数据转换处理，获得预处理后的驾驶行为数据。Specifically, the driving behavior data of the driver is collected through data sensors, and the driving behavior data includes vehicle operation data and driver's action data; the driving behavior data is subjected to data anomaly screening processing, outliers are identified or deleted, and inconsistencies are resolved to clean the data, and abnormal data is removed to obtain the driving behavior data after the data anomaly screening processing; the driving behavior data after the data anomaly screening processing is subjected to redundancy removal processing, and it is checked whether there is duplicate data in the driving behavior data. If there is duplicate data, one or more of the data are deleted. After deleting the duplicate data, the driving behavior data is subjected to missing value processing. For the processing of missing values, the missing values in the data are identified and the type and distribution of the missing values are analyzed. According to the type and distribution of the missing values, an appropriate method is adopted. If there is a missing value, If there are few missing values, the interpolation method is used to fill the missing values. If there are a large number of missing values, the average filling method should be used to fill the missing values. After all the missing values are filled, the data de-redundancy processing is completed to obtain the driving behavior data after de-redundancy processing; the driving behavior data after de-redundancy processing is converted into a readable format, and its unit and time format are unified. After the unification processing is completed, the driving behavior data is normalized, and the maximum-minimum normalization method is used to change the driving behavior data to a new specific area space, and the driving behavior data set is converted into a numerical type according to the numerical value in the new specific area space. The data conversion processing is completed to obtain the pre-processed driving behavior data.
S12：对预处理后的驾驶行为数据进行特征分析，并基于特征分析结果利用预设特征阈值标定各类危险驾驶行为；S12: Performing feature analysis on the preprocessed driving behavior data, and calibrating various types of dangerous driving behaviors using preset feature thresholds based on the feature analysis results;
在本发明具体实施过程中，所述对预处理后的驾驶行为数据进行特征分析，并基于特征分析结果利用预设特征阈值标定各类危险驾驶行为，包括：对预处理后的驾驶行为数据进行分段，获得若干段驾驶行为分段数据；对每一段驾驶行为分段数据计算对应的均值、最大值、最小值和标准差，并基于每一段驾驶行为分段数据对应的均值、最大值、最小值和标准差获得特征指标，所述特征指标包括最大行驶车速、最大行驶加速度、最大航向角速度、注视点位置均值、眨眼频率和瞳孔面积变化率；将所述特征指标与预设特征阈值进行比较，基于比较结果标定各类危险驾驶行为。In the specific implementation process of the present invention, the pre-processed driving behavior data is subjected to feature analysis, and various types of dangerous driving behaviors are calibrated based on the feature analysis results using preset feature thresholds, including: segmenting the pre-processed driving behavior data to obtain a plurality of segments of driving behavior segmented data; calculating the corresponding mean, maximum, minimum and standard deviation for each segment of driving behavior segmented data, and obtaining feature indicators based on the mean, maximum, minimum and standard deviation corresponding to each segment of driving behavior segmented data, wherein the feature indicators include maximum driving speed, maximum driving acceleration, maximum heading angular velocity, gaze point position mean, blinking frequency and pupil area change rate; comparing the feature indicators with preset feature thresholds, and calibrating various types of dangerous driving behaviors based on the comparison results.
进一步的，所述将所述特征指标与预设特征阈值进行比较，基于比较结果标定各类危险驾驶行为，包括若所述最大行驶车速大于预设特征阈值中的最大车速阈值，则标定为超速驾驶；若所述最大行驶加速度大于预设特征阈值中的最大加速度阈值，则标定为急加速驾驶；若所述最大航向角速度大于预设特征阈值中的最大航向角速度阈值，则标定为急变道驾驶；若所述注视点位置均值大于预设特征阈值中的最大注视点偏移距离或所述注视点位置均值小于预设特征阈值中的最小注视点偏移距离，则标定为分心驾驶；若所述眨眼频率小于预设特征阈值中的最小眨眼频率和/或所述瞳孔面积变化率小于预设特征阈值中的最小瞳孔面积变化率，则标定为疲劳驾驶。Furthermore, the characteristic index is compared with a preset characteristic threshold, and various types of dangerous driving behaviors are calibrated based on the comparison results, including: if the maximum driving speed is greater than the maximum speed threshold in the preset characteristic threshold, it is calibrated as speeding; if the maximum driving acceleration is greater than the maximum acceleration threshold in the preset characteristic threshold, it is calibrated as sudden acceleration driving; if the maximum heading angular velocity is greater than the maximum heading angular velocity threshold in the preset characteristic threshold, it is calibrated as sudden lane change driving; if the mean of the gaze point position is greater than the maximum gaze point offset distance in the preset characteristic threshold or the mean of the gaze point position is less than the minimum gaze point offset distance in the preset characteristic threshold, it is calibrated as distracted driving; if the blinking frequency is less than the minimum blinking frequency in the preset characteristic threshold and/or the pupil area change rate is less than the minimum pupil area change rate in the preset characteristic threshold, it is calibrated as fatigue driving.
具体的，对预处理后的驾驶行为数据进行分段，通过指针初始化一个滑动窗口，滑动窗口的大小由现有的驾驶行为规律获取，从预处理的驾驶行为数据中截取固定长度的窗口进行分析，通过已确定大小和长度的滑动窗口进行滑动，对预处理后的驾驶行为数据进行分段，如车辆运行数据中的行驶速度划分为一个分段数据和驾驶员动作数据中的眨眼频率划分为一个分段数据等，获得若干段驾驶行为分段数据；对每一段驾驶行为分段数据计算对应的均值、最大值、最小值和标准差，对每一段驾驶行为分段数据进行比较，获得对应的最大值和最小值，通过驾驶行为分段数据的总和和总数利用滑动窗口的分段值计算对应的驾驶行为分段数据的平均值，对于驾驶行为分段数据中的每一个数据，计算其与平均值的差值，对每个差值进行平方操作，得到差值的平方，计算差值平方的平均值，再对差值平方的平方值进行开根号操作，得到对应的标准差，基于每一段驾驶行为分段数据对应的均值、最大值、最小值和标准差获得特征指标，以驾驶行为分段数据中行驶速度的最大值为特征指标，以驾驶行为分段数据中行驶加速度的最大值为特征指标，以驾驶行为分段数据中航向角速度的最大值为特征指标，以驾驶行为分段数据中注视点位置的平均值为特征指标，以驾驶行为分段数据中眨眼频率的最大值和最小值为特征指标，以瞳孔面积的变化标准差为特征指标，所述特征指标包括最大行驶车速、最大行驶加速度、最大航向角速度、注视点位置均值、眨眼频率和瞳孔面积变化率；将最大行驶车速与最大车速阈值进行比较，若所述最大行驶车速大于预设特征阈值中的最大车速阈值，则标定为超速驾驶；将最大行驶加速度与最大加速度阈值进行比较，若所述最大行驶加速度大于预设特征阈值中的最大加速度阈值，则标定为急加速驾驶；将最大航向角速度与最大航向角速度阈值进行比较，若所述最大航向角速度大于预设特征阈值中的最大航向角速度阈值，则标定为急变道驾驶，还可结合车辆是否在打转向灯的情况下考虑；将注视点位置均值与最大注视点偏移距离和最小注视点偏移距离进行比较，注视点偏移距离分为水平偏移距离和垂直偏移距离，若所述注视点位置均值大于预设特征阈值中的最大注视点偏移距离或所述注视点位置均值小于预设特征阈值中的最小注视点偏移距离，超出了偏移距离则表示驾驶员的注意力不在驾驶道路上，而在其他与驾驶无关的行为上，则标定为分心驾驶；将眨眼频率与最小眨眼频率和瞳孔面积变化率与最小瞳孔面积变化率进行比较，若所述眨眼频率小于预设特征阈值中的最小眨眼频率和/或所述瞳孔面积变化率小于预设特征阈值中的最小瞳孔面积变化率，则标定为疲劳驾驶；根据对驾驶行为数据的特征分析结果利用预设特征阈值标定各类危险行为，能够更准确地将危险驾驶行为进行具体的标定分类。Specifically, the preprocessed driving behavior data is segmented, a sliding window is initialized through a pointer, the size of the sliding window is obtained by the existing driving behavior law, a window of fixed length is intercepted from the preprocessed driving behavior data for analysis, and the preprocessed driving behavior data is segmented by sliding the sliding window of determined size and length, such as dividing the driving speed in the vehicle operation data into a segmented data and the blinking frequency in the driver action data into a segmented data, etc., to obtain several segments of driving behavior segmented data; the corresponding mean, maximum value, minimum value and standard deviation are calculated for each segment of driving behavior segmented data, each segment of driving behavior segmented data is compared to obtain the corresponding maximum value and minimum value, and the corresponding driving behavior segment number is calculated by the segment value of the sliding window through the sum and total of the driving behavior segmented data. The average value of the data is calculated, for each data in the driving behavior segmented data, the difference between the data and the average value is calculated, each difference is squared to obtain the square of the difference, the average value of the square of the difference is calculated, and then the square root of the square of the difference is taken to obtain the corresponding standard deviation. The characteristic index is obtained based on the mean, maximum value, minimum value and standard deviation corresponding to each segment of the driving behavior segmented data. The maximum value of the driving speed in the driving behavior segmented data is used as the characteristic index, the maximum value of the driving acceleration in the driving behavior segmented data is used as the characteristic index, the maximum value of the heading angular velocity in the driving behavior segmented data is used as the characteristic index, the average value of the gaze point position in the driving behavior segmented data is used as the characteristic index, the maximum and minimum values of the blinking frequency in the driving behavior segmented data are used as the characteristic index, and the standard deviation of the change in pupil area is used as the characteristic index. The characteristic index Including maximum driving speed, maximum driving acceleration, maximum heading angular velocity, mean value of gaze point position, blinking frequency and pupil area change rate; compare the maximum driving speed with the maximum speed threshold, if the maximum driving speed is greater than the maximum speed threshold in the preset feature threshold, it is calibrated as speeding; compare the maximum driving acceleration with the maximum acceleration threshold, if the maximum driving acceleration is greater than the maximum acceleration threshold in the preset feature threshold, it is calibrated as sudden acceleration driving; compare the maximum heading angular velocity with the maximum heading angular velocity threshold, if the maximum heading angular velocity is greater than the maximum heading angular velocity threshold in the preset feature threshold, it is calibrated as sudden lane change driving, and it can also be considered in combination with whether the vehicle is turning on the turn signal; compare the mean value of the gaze point position with the maximum gaze point offset distance and the minimum gaze point offset distance, the gaze point offset The distance is divided into horizontal offset distance and vertical offset distance. If the mean value of the gaze point position is greater than the maximum gaze point offset distance in the preset feature threshold or the mean value of the gaze point position is less than the minimum gaze point offset distance in the preset feature threshold, exceeding the offset distance indicates that the driver's attention is not on the driving road, but on other behaviors not related to driving, and it is calibrated as distracted driving; the blinking frequency is compared with the minimum blinking frequency and the pupil area change rate is compared with the minimum pupil area change rate. If the blinking frequency is less than the minimum blinking frequency in the preset feature threshold and/or the pupil area change rate is less than the minimum pupil area change rate in the preset feature threshold, it is calibrated as fatigue driving; various types of dangerous behaviors are calibrated using preset feature thresholds based on the feature analysis results of the driving behavior data, so that dangerous driving behaviors can be more accurately calibrated and classified.
S13：基于各类危险驾驶行为的标定结果利用聚类分析法进行聚类，获得聚类结果；S13: clustering based on the calibration results of various dangerous driving behaviors using a cluster analysis method to obtain clustering results;
在本发明具体实施过程中，所述基于各类危险驾驶行为的标定结果利用聚类分析法进行聚类，获得聚类结果，包括：计算每类危险驾驶行为的标定结果超出预设特征阈值的幅值，利用聚类分析法对所有超出预设特征阈值的幅值进行聚类，获得各类危险驾驶行为的分类占比。In the specific implementation process of the present invention, the calibration results based on various types of dangerous driving behaviors are clustered using a cluster analysis method to obtain clustering results, including: calculating the amplitude of the calibration results of each type of dangerous driving behavior that exceeds a preset feature threshold, and using a cluster analysis method to cluster all amplitudes that exceed the preset feature threshold to obtain the classification ratio of each type of dangerous driving behavior.
具体的，当标定出驾驶员的危险驾驶行为后，计算其危险驾驶行为的标定结果超出预设特征阈值的幅值，如标定出驾驶员发生超速驾驶，则计算其所标定的最大行驶车速与最大车速阈值的差值，若标定出驾驶员发生分心驾驶，则计算其所标定的注视点位置均值与最大注视点偏移距离或最小注视点偏移距离的差值；利用聚类分析法对所有超出预设特征阈值的幅值进行聚类，对所有超出预设特征阈值的幅值选择严重性变量，采用谱聚类的方法根据严重性变量形成各类危险驾驶行为的分类占比。Specifically, after the driver's dangerous driving behavior is calibrated, the amplitude of the calibration result of the dangerous driving behavior exceeding the preset feature threshold is calculated. For example, if the driver is calibrated to be speeding, the difference between the calibrated maximum driving speed and the maximum speed threshold is calculated. If the driver is calibrated to be distracted driving, the difference between the calibrated gaze point position mean and the maximum gaze point offset distance or the minimum gaze point offset distance is calculated. The cluster analysis method is used to cluster all amplitudes exceeding the preset feature threshold, and the severity variable is selected for all amplitudes exceeding the preset feature threshold. The spectral clustering method is used to form the classification proportion of each type of dangerous driving behavior according to the severity variable.
S14：基于聚类结果计算类别权重，基于所述类别权重量化各类危险驾驶行为的相对风险，并基于所述类别权重和相对风险计算驾驶行为得分；S14: calculating a category weight based on the clustering result, quantifying the relative risk of each category of dangerous driving behavior based on the category weight, and calculating a driving behavior score based on the category weight and the relative risk;
在本发明具体实施过程中，所述基于聚类结果计算类别权重，基于所述类别权重量化各类危险驾驶行为的相对风险，并基于所述类别权重和相对风险计算驾驶行为得分，包括：基于聚类结果计算各类危险驾驶行为对应的初始评价指标；对对应的初始评价指标中的极小型指标进行正向化处理，获得评价正向化指标，并对所述评价正向化指标进行标准化处理，获得标准化正向矩阵；基于所述标准化正向矩阵利用熵权法计算对应的类别权重；基于对应的类别权重建立在预设单位时间中各类危险驾驶行为的单位距离风险，基于所述单位距离风险量化各类危险驾驶行为的相对风险；对各类危险驾驶行为对应的类别权重和相对风险进行变量替换，基于变量替换后的类别权重和相对风险计算驾驶行为得分。In the specific implementation process of the present invention, the category weights are calculated based on the clustering results, the relative risks of various types of dangerous driving behaviors are quantified based on the category weights, and the driving behavior scores are calculated based on the category weights and relative risks, including: calculating the initial evaluation indicators corresponding to various types of dangerous driving behaviors based on the clustering results; performing positive processing on the extremely small indicators in the corresponding initial evaluation indicators to obtain positive evaluation indicators, and performing standardization processing on the positive evaluation indicators to obtain a standardized forward matrix; calculating the corresponding category weights based on the standardized forward matrix using the entropy weight method; establishing the unit distance risks of various types of dangerous driving behaviors in a preset unit time based on the corresponding category weights, and quantifying the relative risks of various types of dangerous driving behaviors based on the unit distance risks; performing variable replacement on the category weights and relative risks corresponding to various types of dangerous driving behaviors, and calculating the driving behavior scores based on the category weights and relative risks after variable replacement.
进一步的，所述基于所述单位距离风险量化各类危险驾驶行为的相对风险，包括：基于所述单位距离风险利用交叉指数法构建交叉指数矩阵；基于所述交叉指数矩阵量化各类危险驾驶行为的相对风险。Furthermore, the quantifying the relative risks of various dangerous driving behaviors based on the unit distance risk includes: constructing a cross-index matrix based on the unit distance risk using a cross-index method; and quantifying the relative risks of various dangerous driving behaviors based on the cross-index matrix.
具体的，基于聚类结果的分类占比计算各类危险驾驶行为对应的初始评价指标，通过分类占比初步确定其评价指标；对对应的初始评价指标中的极小型指标进行正向化处理，将极小型指标转换为极大型指标，得到对应驾驶员的评价正向化指标；对所述评价正向化指标进行标准化处理，为了排除不同量纲的影响对正向化矩阵进行标准化处理，获得标准化正向矩阵；基于所述标准化正向矩阵利用熵权法计算对应的类别权重，根据标准化正向矩阵计算各评价指标的信息熵，根据信息熵可说明指标数据分布情况，根据信息熵计算对应的类别权重，信息熵是处理多维数据常用的一种方法，信息熵越大，表示组内系统越混乱，包含的信息也就越丰富，相对的来说这个指标就越有意义，类别权重的计算表达式为：Specifically, the initial evaluation index corresponding to each type of dangerous driving behavior is calculated based on the classification proportion of the clustering result, and the evaluation index is preliminarily determined by the classification proportion; the extremely small index in the corresponding initial evaluation index is positively processed, and the extremely small index is converted into an extremely large index to obtain the evaluation positive index of the corresponding driver; the evaluation positive index is standardized, and the positive matrix is standardized to eliminate the influence of different dimensions to obtain a standardized forward matrix; based on the standardized forward matrix, the corresponding category weight is calculated using the entropy weight method, and the information entropy of each evaluation index is calculated according to the standardized forward matrix. The distribution of the indicator data can be explained according to the information entropy, and the corresponding category weight is calculated according to the information entropy. Information entropy is a commonly used method for processing multidimensional data. The larger the information entropy, the more chaotic the system in the group is, and the richer the information contained is. Relatively speaking, this indicator is more meaningful. The calculation expression of the category weight is:
其中，Wj为类别权重，指标的信息熵分别为E1,E2,…,En；对得到的类别权重进行检验，确保其合理性和可靠性，检验完成后便可得到对应的类别权重；基于对应的类别权重建立在预设单位时间中各类危险驾驶行为的单位距离风险，采用面积法，基于对应的类别权重建立单位时间内超过限速值的速度所对应的面积与其对应速度风险之间的联系，将所有行驶时间内的速度风险进行累加求和，即可得到针对超速危险驾驶行为的路段驾驶总风险，并将其除以路段距离得到其对应的单位距离风险，各类危险驾驶行为的单位距离风险都是按上述处理，直至将所有危险驾驶行为的单位距离风险都得到为止；基于所述单位距离风险利用交叉指数法构建交叉指数矩阵，结合单位距离风险通过expm函数构建交叉指数矩阵，expm函数是一种用于计算指数矩阵的函数；基于所述交叉指数矩阵量化各类危险驾驶行为的相对风险；对各类危险驾驶行为对应的类别权重和相对风险进行变量替换，变量替换是为了处理类别权重和相对风险计算中的非线性问题，基于变量替换后的类别权重和相对风险计算驾驶行为得分，通过计算类别权重和危险驾驶行为的相对风险来计算驾驶行为得分，能够更为客观地评定计算驾驶员的驾驶行为得分。Among them,Wj is the category weight, and the information entropy of the indicators isE1 ,E2 , …,En respectively; the obtained category weight is tested to ensure its rationality and reliability, and the corresponding category weight can be obtained after the test is completed; based on the corresponding category weight, the unit distance risk of various dangerous driving behaviors in the preset unit time is established, and the area method is used to establish the relationship between the area corresponding to the speed exceeding the speed limit value in the unit time and its corresponding speed risk based on the corresponding category weight. The speed risks in all driving times are accumulated and summed to obtain the total road section driving risk for speeding dangerous driving behaviors, and it is divided by the road section distance to obtain its corresponding unit distance risk. The unit distance risks of various dangerous driving behaviors are processed as above until the unit distance risks of all dangerous driving behaviors are obtained. To the end; construct a cross-index matrix based on the unit distance risk using the cross-index method, and construct a cross-index matrix through the expm function in combination with the unit distance risk, the expm function is a function used to calculate the index matrix; quantify the relative risks of various types of dangerous driving behaviors based on the cross-index matrix; perform variable replacement on the category weights and relative risks corresponding to various types of dangerous driving behaviors, the variable replacement is to deal with the nonlinear problems in the calculation of category weights and relative risks, calculate the driving behavior score based on the category weights and relative risks after variable replacement, and calculate the driving behavior score by calculating the category weights and the relative risks of dangerous driving behaviors, which can more objectively evaluate and calculate the driver's driving behavior score.
S15：获取驾驶员的心理素质数据，基于所述心理素质数据利用预设要素测评值计算驾驶员的心理状况得分，并基于所述心理状况得分和驾驶行为得分计算驾驶员的综合风险得分；S15: Acquire the driver's psychological quality data, calculate the driver's psychological condition score based on the psychological quality data using the preset factor evaluation value, and calculate the driver's comprehensive risk score based on the psychological condition score and the driving behavior score;
在本发明具体实施过程中，所述获取驾驶员的心理素质数据，基于所述心理素质数据利用预设要素测评值计算驾驶员的心理状况得分，并基于所述心理状况得分和驾驶行为得分计算驾驶员的综合风险得分，包括：采集驾驶员的心理素质数据，所述心理素质数据包括心理认知能力、人格特质和心理健康状况；对驾驶员的心理认知能力、人格特质和心理健康状况分别赋予对应的加权权重，基于对应的加权权重利用预设要素测评值计算驾驶员的心理状况得分；基于所述心理状况得分和驾驶行为得分利用预设得分比例计算驾驶员的综合风险得分。In the specific implementation process of the present invention, the driver's psychological quality data is obtained, the driver's psychological condition score is calculated based on the psychological quality data using preset factor evaluation values, and the driver's comprehensive risk score is calculated based on the psychological condition score and the driving behavior score, including: collecting the driver's psychological quality data, the psychological quality data includes psychological cognitive ability, personality traits and mental health status; assigning corresponding weighted weights to the driver's psychological cognitive ability, personality traits and mental health status, respectively, and calculating the driver's psychological condition score based on the corresponding weighted weights using preset factor evaluation values; calculating the driver's comprehensive risk score based on the psychological condition score and the driving behavior score using a preset score ratio.
具体的，通过对接对应的管理平台采集驾驶员的心理素质数据，心理素质数据包括心理认知能力、人格特质和心理健康状况；对驾驶员的心理认知能力、人格特质和心理健康状况分别赋予对应的加权权重，根据预设标注你对每项数据都赋予不同的加权权重，通过预设要素测评值对心理认知能力、人格特质和心理健康状况对应的加权权重计算驾驶员的心理状况得分，通过要素测评值能够从不同的维度来评定驾驶员的心理状况得分，基于所述心理状况得分和驾驶行为得分利用预设得分比例计算驾驶员的综合风险得分，心理状况得分和驾驶行为得分分别与对应的预设比例相乘，将两者相乘的结果再相加，得到驾驶员的综合风险得分，将驾驶员的心理状况得分也加入至风险得分的考虑中，通过由心理状况得分和驾驶行为得分所生成的综合风险得分能够更为准确且客观地对驾驶员进行风险评价。Specifically, the psychological quality data of the driver is collected by connecting to the corresponding management platform, and the psychological quality data includes psychological cognitive ability, personality traits and mental health status; corresponding weighted weights are assigned to the driver's psychological cognitive ability, personality traits and mental health status respectively, and different weighted weights are assigned to each data according to the preset annotations, and the driver's psychological condition score is calculated by the weighted weights corresponding to the psychological cognitive ability, personality traits and mental health status through the preset factor evaluation values. The driver's psychological condition score can be evaluated from different dimensions through the factor evaluation values, and the driver's comprehensive risk score is calculated based on the psychological condition score and the driving behavior score using the preset score ratio. The psychological condition score and the driving behavior score are respectively multiplied by the corresponding preset ratios, and the results of the multiplication of the two are added together to obtain the driver's comprehensive risk score, and the driver's psychological condition score is also added into the risk score consideration, and the comprehensive risk score generated by the psychological condition score and the driving behavior score can more accurately and objectively evaluate the risk of the driver.
S16：基于驾驶员的综合风险得分对驾驶员进行风险等级评级，基于驾驶员的风险等级生成对应的管控策略，并将管控策略发送至对应的驾驶员。S16: Rating the driver's risk level based on the driver's comprehensive risk score, generating a corresponding control strategy based on the driver's risk level, and sending the control strategy to the corresponding driver.
在本发明具体实施过程中，所述基于驾驶员的综合风险得分对驾驶员进行风险等级评级，基于驾驶员的风险等级生成对应的管控策略，并将管控策略发送至对应的驾驶员，包括：基于驾驶员的综合风险得分利用预设评级标准对驾驶员划分风险等级；基于驾驶员的风险等级利用对驾驶员所标定的各类危险驾驶行为生成对应的管控策略，并将对应的管控策略实时发送至对应的驾驶员中。During the specific implementation of the present invention, the driver's risk level is rated based on the driver's comprehensive risk score, a corresponding management and control strategy is generated based on the driver's risk level, and the management and control strategy is sent to the corresponding driver, including: dividing the driver's risk level based on the driver's comprehensive risk score using preset rating standards; generating corresponding management and control strategies based on the driver's risk level using various types of dangerous driving behaviors calibrated for the driver, and sending the corresponding management and control strategies to the corresponding driver in real time.
具体的，基于驾驶员的综合风险得分利用预设评级标准对驾驶员划分风险等级，根据不同的综合风险得分分别划分为低风险等级、中低风险等级、中风险等级、中高风险等级和高风险；基于驾驶员的风险等级利用对驾驶员所标定的各类危险驾驶行为生成对应的管控策略，对驾驶员做出相应的具体管理限制和行为更正，如驾驶员的风险等级为高风险，则需要对应的驾驶员每隔一个时段便进行汇报，根据其所发生的危险驾驶行为进行更正，直至驾驶员的风险等级降低至预设标准，才解除对驾驶员的相应管理限制。Specifically, based on the comprehensive risk score of the driver, the driver is divided into risk levels using preset rating standards, and is divided into low risk level, medium-low risk level, medium risk level, medium-high risk level and high risk according to different comprehensive risk scores; based on the driver's risk level, a corresponding management and control strategy is generated for each type of dangerous driving behavior calibrated for the driver, and corresponding specific management restrictions and behavioral corrections are made to the driver. If the driver's risk level is high risk, the corresponding driver is required to report at regular intervals and make corrections based on the dangerous driving behavior that has occurred until the driver's risk level is reduced to the preset standard, and then the corresponding management restrictions on the driver are lifted.
在本发明实施例中，根据对驾驶行为数据的特征分析结果利用预设特征阈值标定各类危险行为，将危险驾驶行为进行具体的标定分类，再通过计算类别权重和危险驾驶行为的相对风险来计算驾驶行为得分，能够更为客观地评定计算驾驶员的驾驶行为得分，并将驾驶员的心理状况得分也加入至风险得分的考虑中，通过由心理状况得分和驾驶行为得分所生成的综合风险得分能够更为准确且客观地对驾驶员进行风险评价，从而能够更为准确且具体地选取对应的管控策略，对驾驶员的管理能够达到更好的效果。In an embodiment of the present invention, various types of dangerous behaviors are calibrated using preset characteristic thresholds based on the characteristic analysis results of driving behavior data, and the dangerous driving behaviors are specifically calibrated and classified. Then, the driving behavior score is calculated by calculating the category weight and the relative risk of the dangerous driving behavior. The driving behavior score of the driver can be evaluated and calculated more objectively, and the driver's psychological condition score is also added to the risk score consideration. The comprehensive risk score generated by the psychological condition score and the driving behavior score can more accurately and objectively evaluate the risk of the driver, so that the corresponding control strategy can be selected more accurately and specifically, and the management of the driver can achieve better results.
实施例二Embodiment 2
请参阅图2，图1是本发明实施例中的驾驶员风险评价管理系统的结构组成示意图。Please refer to FIG. 2 , FIG. 1 is a schematic diagram of the structural composition of a driver risk assessment management system in an embodiment of the present invention.
如图2所示，一种驾驶员风险评价管理系统，所述系统包括：As shown in FIG2 , a driver risk assessment management system includes:
数据预处理模块21：用于获取驾驶员的驾驶行为数据，并对驾驶员的驾驶行为数据进行数据预处理，获得预处理后的驾驶行为数据；Data preprocessing module 21: used to obtain the driving behavior data of the driver, and perform data preprocessing on the driving behavior data of the driver to obtain the preprocessed driving behavior data;
在本发明具体实施过程中，所述获取驾驶员的驾驶行为数据，并对驾驶员的驾驶行为数据进行数据预处理，获得预处理后的驾驶行为数据，包括：基于数据传感器采集驾驶员的驾驶行为数据；对所述驾驶行为数据进行数据异常筛除处理，获得数据异常筛除处理后的驾驶行为数据；对数据异常筛除处理后的驾驶行为数据进行去冗处理，获得去冗处理后的驾驶行为数据；对去冗处理后的驾驶行为数据进行数据转换处理，获得预处理后的驾驶行为数据。In the specific implementation process of the present invention, the method of acquiring the driving behavior data of the driver and performing data preprocessing on the driving behavior data of the driver to obtain the preprocessed driving behavior data includes: collecting the driving behavior data of the driver based on a data sensor; performing data anomaly screening processing on the driving behavior data to obtain the driving behavior data after the data anomaly screening processing; performing redundancy processing on the driving behavior data after the data anomaly screening processing to obtain the driving behavior data after the redundancy processing; performing data conversion processing on the driving behavior data after the redundancy processing to obtain the preprocessed driving behavior data.
具体的，通过数据传感器采集驾驶员的驾驶行为数据，驾驶行为数据包括车辆运行数据和驾驶员的动作数据；对驾驶行为数据进行数据异常筛除处理，识别或删除离群点并解决不一致性来清理数据，清除异常数据，获得数据异常筛除处理后的驾驶行为数据；对数据异常筛除处理后的驾驶行为数据进行去冗处理，检查驾驶行为数据中是否存在重复的数据，若有重复数据，则删除其中的一条或多条数据，删除重复数据后便对驾驶行为数据进行缺失值的处理，对于数据缺失值的处理，识别数据中存在的缺失值并分析缺失值的类型和分布情况，根据缺失值的类型和分布情况采用适当的方法，若缺失值较少，则采用插值方法进行填补缺失值，若存在大量缺失值，则应采用平均数填补法对缺失值进行填补，直至将所有缺失值填补完成后便完成数据去冗处理，获得去冗处理后的驾驶行为数据；对去冗处理后的驾驶行为数据进行数据转换处理，将去冗处理后的驾驶行为数据转换成可读取的格式，并将其单位和时间格式进行统一化处理，完成统一化处理后，便对驾驶行为数据进行数据规范化处理，采用极大值-极小值规范化方法，将驾驶行为数据变化到新的特定区域空间中，按新特定区域空间中的数值将驾驶行为数据集转换数值类型，完成数据转换处理，获得预处理后的驾驶行为数据。Specifically, the driving behavior data of the driver is collected through data sensors, and the driving behavior data includes vehicle operation data and driver's action data; the driving behavior data is subjected to data anomaly screening processing, outliers are identified or deleted, and inconsistencies are resolved to clean the data, and abnormal data is removed to obtain the driving behavior data after the data anomaly screening processing; the driving behavior data after the data anomaly screening processing is subjected to redundancy removal processing, and it is checked whether there is duplicate data in the driving behavior data. If there is duplicate data, one or more of the data are deleted. After deleting the duplicate data, the driving behavior data is subjected to missing value processing. For the processing of missing values, the missing values in the data are identified and the type and distribution of the missing values are analyzed. According to the type and distribution of the missing values, an appropriate method is adopted. If there is a missing value, If there are few missing values, the interpolation method is used to fill the missing values. If there are a large number of missing values, the average filling method should be used to fill the missing values. After all the missing values are filled, the data de-redundancy processing is completed to obtain the driving behavior data after de-redundancy processing; the driving behavior data after de-redundancy processing is converted into a readable format, and its unit and time format are unified. After the unification processing is completed, the driving behavior data is normalized, and the maximum-minimum normalization method is used to change the driving behavior data to a new specific area space, and the driving behavior data set is converted into a numerical type according to the numerical value in the new specific area space. The data conversion processing is completed to obtain the pre-processed driving behavior data.
危险驾驶行为标定模块22：用于对预处理后的驾驶行为数据进行特征分析，并基于特征分析结果利用预设特征阈值标定各类危险驾驶行为；Dangerous driving behavior calibration module 22: used to perform feature analysis on the pre-processed driving behavior data, and calibrate various types of dangerous driving behaviors using preset feature thresholds based on the feature analysis results;
在本发明具体实施过程中，所述对预处理后的驾驶行为数据进行特征分析，并基于特征分析结果利用预设特征阈值标定各类危险驾驶行为，包括：对预处理后的驾驶行为数据进行分段，获得若干段驾驶行为分段数据；对每一段驾驶行为分段数据计算对应的均值、最大值、最小值和标准差，并基于每一段驾驶行为分段数据对应的均值、最大值、最小值和标准差获得特征指标，所述特征指标包括最大行驶车速、最大行驶加速度、最大航向角速度、注视点位置均值、眨眼频率和瞳孔面积变化率；将所述特征指标与预设特征阈值进行比较，基于比较结果标定各类危险驾驶行为。In the specific implementation process of the present invention, the pre-processed driving behavior data is subjected to feature analysis, and various types of dangerous driving behaviors are calibrated based on the feature analysis results using preset feature thresholds, including: segmenting the pre-processed driving behavior data to obtain a plurality of segments of driving behavior segmented data; calculating the corresponding mean, maximum, minimum and standard deviation for each segment of driving behavior segmented data, and obtaining feature indicators based on the mean, maximum, minimum and standard deviation corresponding to each segment of driving behavior segmented data, wherein the feature indicators include maximum driving speed, maximum driving acceleration, maximum heading angular velocity, gaze point position mean, blinking frequency and pupil area change rate; comparing the feature indicators with preset feature thresholds, and calibrating various types of dangerous driving behaviors based on the comparison results.
进一步的，所述将所述特征指标与预设特征阈值进行比较，基于比较结果标定各类危险驾驶行为，包括若所述最大行驶车速大于预设特征阈值中的最大车速阈值，则标定为超速驾驶；若所述最大行驶加速度大于预设特征阈值中的最大加速度阈值，则标定为急加速驾驶；若所述最大航向角速度大于预设特征阈值中的最大航向角速度阈值，则标定为急变道驾驶；若所述注视点位置均值大于预设特征阈值中的最大注视点偏移距离或所述注视点位置均值小于预设特征阈值中的最小注视点偏移距离，则标定为分心驾驶；若所述眨眼频率小于预设特征阈值中的最小眨眼频率和/或所述瞳孔面积变化率小于预设特征阈值中的最小瞳孔面积变化率，则标定为疲劳驾驶。Furthermore, the characteristic index is compared with a preset characteristic threshold, and various types of dangerous driving behaviors are calibrated based on the comparison results, including: if the maximum driving speed is greater than the maximum speed threshold in the preset characteristic threshold, it is calibrated as speeding; if the maximum driving acceleration is greater than the maximum acceleration threshold in the preset characteristic threshold, it is calibrated as sudden acceleration driving; if the maximum heading angular velocity is greater than the maximum heading angular velocity threshold in the preset characteristic threshold, it is calibrated as sudden lane change driving; if the mean of the gaze point position is greater than the maximum gaze point offset distance in the preset characteristic threshold or the mean of the gaze point position is less than the minimum gaze point offset distance in the preset characteristic threshold, it is calibrated as distracted driving; if the blinking frequency is less than the minimum blinking frequency in the preset characteristic threshold and/or the pupil area change rate is less than the minimum pupil area change rate in the preset characteristic threshold, it is calibrated as fatigue driving.
具体的，对预处理后的驾驶行为数据进行分段，通过指针初始化一个滑动窗口，滑动窗口的大小由现有的驾驶行为规律获取，从预处理的驾驶行为数据中截取固定长度的窗口进行分析，通过已确定大小和长度的滑动窗口进行滑动，对预处理后的驾驶行为数据进行分段，如车辆运行数据中的行驶速度划分为一个分段数据和驾驶员动作数据中的眨眼频率划分为一个分段数据等，获得若干段驾驶行为分段数据；对每一段驾驶行为分段数据计算对应的均值、最大值、最小值和标准差，对每一段驾驶行为分段数据进行比较，获得对应的最大值和最小值，通过驾驶行为分段数据的总和和总数利用滑动窗口的分段值计算对应的驾驶行为分段数据的平均值，对于驾驶行为分段数据中的每一个数据，计算其与平均值的差值，对每个差值进行平方操作，得到差值的平方，计算差值平方的平均值，再对差值平方的平方值进行开根号操作，得到对应的标准差，基于每一段驾驶行为分段数据对应的均值、最大值、最小值和标准差获得特征指标，以驾驶行为分段数据中行驶速度的最大值为特征指标，以驾驶行为分段数据中行驶加速度的最大值为特征指标，以驾驶行为分段数据中航向角速度的最大值为特征指标，以驾驶行为分段数据中注视点位置的平均值为特征指标，以驾驶行为分段数据中眨眼频率的最大值和最小值为特征指标，以瞳孔面积的变化标准差为特征指标，所述特征指标包括最大行驶车速、最大行驶加速度、最大航向角速度、注视点位置均值、眨眼频率和瞳孔面积变化率；将最大行驶车速与最大车速阈值进行比较，若所述最大行驶车速大于预设特征阈值中的最大车速阈值，则标定为超速驾驶；将最大行驶加速度与最大加速度阈值进行比较，若所述最大行驶加速度大于预设特征阈值中的最大加速度阈值，则标定为急加速驾驶；将最大航向角速度与最大航向角速度阈值进行比较，若所述最大航向角速度大于预设特征阈值中的最大航向角速度阈值，则标定为急变道驾驶，还可结合车辆是否在打转向灯的情况下考虑；将注视点位置均值与最大注视点偏移距离和最小注视点偏移距离进行比较，注视点偏移距离分为水平偏移距离和垂直偏移距离，若所述注视点位置均值大于预设特征阈值中的最大注视点偏移距离或所述注视点位置均值小于预设特征阈值中的最小注视点偏移距离，超出了偏移距离则表示驾驶员的注意力不在驾驶道路上，而在其他与驾驶无关的行为上，则标定为分心驾驶；将眨眼频率与最小眨眼频率和瞳孔面积变化率与最小瞳孔面积变化率进行比较，若所述眨眼频率小于预设特征阈值中的最小眨眼频率和/或所述瞳孔面积变化率小于预设特征阈值中的最小瞳孔面积变化率，则标定为疲劳驾驶；根据对驾驶行为数据的特征分析结果利用预设特征阈值标定各类危险行为，能够更准确地将危险驾驶行为进行具体的标定分类。Specifically, the preprocessed driving behavior data is segmented, a sliding window is initialized through a pointer, the size of the sliding window is obtained by the existing driving behavior law, a window of fixed length is intercepted from the preprocessed driving behavior data for analysis, and the preprocessed driving behavior data is segmented by sliding the sliding window of determined size and length, such as dividing the driving speed in the vehicle operation data into a segmented data and the blinking frequency in the driver action data into a segmented data, etc., to obtain several segments of driving behavior segmented data; the corresponding mean, maximum value, minimum value and standard deviation are calculated for each segment of driving behavior segmented data, each segment of driving behavior segmented data is compared to obtain the corresponding maximum value and minimum value, and the corresponding driving behavior segment number is calculated by the segment value of the sliding window through the sum and total of the driving behavior segmented data. The average value of the data is calculated, for each data in the driving behavior segmented data, the difference between the data and the average value is calculated, each difference is squared to obtain the square of the difference, the average value of the square of the difference is calculated, and then the square root of the square of the difference is taken to obtain the corresponding standard deviation. The characteristic index is obtained based on the mean, maximum value, minimum value and standard deviation corresponding to each segment of the driving behavior segmented data. The maximum value of the driving speed in the driving behavior segmented data is used as the characteristic index, the maximum value of the driving acceleration in the driving behavior segmented data is used as the characteristic index, the maximum value of the heading angular velocity in the driving behavior segmented data is used as the characteristic index, the average value of the gaze point position in the driving behavior segmented data is used as the characteristic index, the maximum and minimum values of the blinking frequency in the driving behavior segmented data are used as the characteristic index, and the standard deviation of the change in pupil area is used as the characteristic index. The characteristic index Including maximum driving speed, maximum driving acceleration, maximum heading angular velocity, mean value of gaze point position, blinking frequency and pupil area change rate; compare the maximum driving speed with the maximum speed threshold, if the maximum driving speed is greater than the maximum speed threshold in the preset feature threshold, it is calibrated as speeding; compare the maximum driving acceleration with the maximum acceleration threshold, if the maximum driving acceleration is greater than the maximum acceleration threshold in the preset feature threshold, it is calibrated as sudden acceleration driving; compare the maximum heading angular velocity with the maximum heading angular velocity threshold, if the maximum heading angular velocity is greater than the maximum heading angular velocity threshold in the preset feature threshold, it is calibrated as sudden lane change driving, and it can also be considered in combination with whether the vehicle is turning on the turn signal; compare the mean value of the gaze point position with the maximum gaze point offset distance and the minimum gaze point offset distance, the gaze point offset The distance is divided into horizontal offset distance and vertical offset distance. If the mean value of the gaze point position is greater than the maximum gaze point offset distance in the preset feature threshold or the mean value of the gaze point position is less than the minimum gaze point offset distance in the preset feature threshold, exceeding the offset distance indicates that the driver's attention is not on the driving road, but on other behaviors not related to driving, and it is calibrated as distracted driving; the blinking frequency is compared with the minimum blinking frequency and the pupil area change rate is compared with the minimum pupil area change rate. If the blinking frequency is less than the minimum blinking frequency in the preset feature threshold and/or the pupil area change rate is less than the minimum pupil area change rate in the preset feature threshold, it is calibrated as fatigue driving; various types of dangerous behaviors are calibrated using preset feature thresholds based on the feature analysis results of the driving behavior data, so that dangerous driving behaviors can be more accurately calibrated and classified.
聚类模块23：用于基于各类危险驾驶行为的标定结果利用聚类分析法进行聚类，获得聚类结果；Clustering module 23: used to perform clustering based on the calibration results of various dangerous driving behaviors using a clustering analysis method to obtain clustering results;
在本发明具体实施过程中，所述基于各类危险驾驶行为的标定结果利用聚类分析法进行聚类，获得聚类结果，包括：计算每类危险驾驶行为的标定结果超出预设特征阈值的幅值，利用聚类分析法对所有超出预设特征阈值的幅值进行聚类，获得各类危险驾驶行为的分类占比。In the specific implementation process of the present invention, the calibration results based on various types of dangerous driving behaviors are clustered using a cluster analysis method to obtain clustering results, including: calculating the amplitude of the calibration results of each type of dangerous driving behavior that exceeds a preset feature threshold, and using a cluster analysis method to cluster all amplitudes that exceed the preset feature threshold to obtain the classification ratio of each type of dangerous driving behavior.
具体的，当标定出驾驶员的危险驾驶行为后，计算其危险驾驶行为的标定结果超出预设特征阈值的幅值，如标定出驾驶员发生超速驾驶，则计算其所标定的最大行驶车速与最大车速阈值的差值，若标定出驾驶员发生分心驾驶，则计算其所标定的注视点位置均值与最大注视点偏移距离或最小注视点偏移距离的差值；利用聚类分析法对所有超出预设特征阈值的幅值进行聚类，对所有超出预设特征阈值的幅值选择严重性变量，采用谱聚类的方法根据严重性变量形成各类危险驾驶行为的分类占比。Specifically, after the driver's dangerous driving behavior is calibrated, the amplitude of the calibration result of the dangerous driving behavior exceeding the preset feature threshold is calculated. For example, if the driver is calibrated to be speeding, the difference between the calibrated maximum driving speed and the maximum speed threshold is calculated. If the driver is calibrated to be distracted driving, the difference between the calibrated gaze point position mean and the maximum gaze point offset distance or the minimum gaze point offset distance is calculated. The cluster analysis method is used to cluster all amplitudes exceeding the preset feature threshold, and the severity variable is selected for all amplitudes exceeding the preset feature threshold. The spectral clustering method is used to form the classification proportion of each type of dangerous driving behavior according to the severity variable.
驾驶行为得分计算模块24：用于基于聚类结果计算类别权重，基于所述类别权重量化各类危险驾驶行为的相对风险，并基于所述类别权重和相对风险计算驾驶行为得分；A driving behavior score calculation module 24 is used to calculate the category weights based on the clustering results, quantify the relative risks of various types of dangerous driving behaviors based on the category weights, and calculate the driving behavior scores based on the category weights and the relative risks;
在本发明具体实施过程中，所述基于聚类结果计算类别权重，基于所述类别权重量化各类危险驾驶行为的相对风险，并基于所述类别权重和相对风险计算驾驶行为得分，包括：基于聚类结果计算各类危险驾驶行为对应的初始评价指标；对对应的初始评价指标中的极小型指标进行正向化处理，获得评价正向化指标，并对所述评价正向化指标进行标准化处理，获得标准化正向矩阵；基于所述标准化正向矩阵利用熵权法计算对应的类别权重；基于对应的类别权重建立在预设单位时间中各类危险驾驶行为的单位距离风险，基于所述单位距离风险量化各类危险驾驶行为的相对风险；对各类危险驾驶行为对应的类别权重和相对风险进行变量替换，基于变量替换后的类别权重和相对风险计算驾驶行为得分。In the specific implementation process of the present invention, the category weights are calculated based on the clustering results, the relative risks of various types of dangerous driving behaviors are quantified based on the category weights, and the driving behavior scores are calculated based on the category weights and relative risks, including: calculating the initial evaluation indicators corresponding to various types of dangerous driving behaviors based on the clustering results; performing positive processing on the extremely small indicators in the corresponding initial evaluation indicators to obtain positive evaluation indicators, and performing standardization processing on the positive evaluation indicators to obtain a standardized forward matrix; calculating the corresponding category weights based on the standardized forward matrix using the entropy weight method; establishing the unit distance risks of various types of dangerous driving behaviors in a preset unit time based on the corresponding category weights, and quantifying the relative risks of various types of dangerous driving behaviors based on the unit distance risks; performing variable replacement on the category weights and relative risks corresponding to various types of dangerous driving behaviors, and calculating the driving behavior scores based on the category weights and relative risks after variable replacement.
进一步的，所述基于所述单位距离风险量化各类危险驾驶行为的相对风险，包括：基于所述单位距离风险利用交叉指数法构建交叉指数矩阵；基于所述交叉指数矩阵量化各类危险驾驶行为的相对风险。Furthermore, the quantifying the relative risks of various dangerous driving behaviors based on the unit distance risk includes: constructing a cross-index matrix based on the unit distance risk using a cross-index method; and quantifying the relative risks of various dangerous driving behaviors based on the cross-index matrix.
具体的，基于聚类结果的分类占比计算各类危险驾驶行为对应的初始评价指标，通过分类占比初步确定其评价指标；对对应的初始评价指标中的极小型指标进行正向化处理，将极小型指标转换为极大型指标，得到对应驾驶员的评价正向化指标；对所述评价正向化指标进行标准化处理，为了排除不同量纲的影响对正向化矩阵进行标准化处理，获得标准化正向矩阵；基于所述标准化正向矩阵利用熵权法计算对应的类别权重，根据标准化正向矩阵计算各评价指标的信息熵，根据信息熵可说明指标数据分布情况，根据信息熵计算对应的类别权重，信息熵是处理多维数据常用的一种方法，信息熵越大，表示组内系统越混乱，包含的信息也就越丰富，相对的来说这个指标就越有意义，类别权重的计算表达式为：Specifically, the initial evaluation index corresponding to each type of dangerous driving behavior is calculated based on the classification proportion of the clustering result, and the evaluation index is preliminarily determined by the classification proportion; the extremely small index in the corresponding initial evaluation index is positively processed, and the extremely small index is converted into an extremely large index to obtain the evaluation positive index of the corresponding driver; the evaluation positive index is standardized, and the positive matrix is standardized to eliminate the influence of different dimensions to obtain a standardized forward matrix; the corresponding category weight is calculated based on the standardized forward matrix using the entropy weight method, and the information entropy of each evaluation index is calculated according to the standardized forward matrix. The distribution of the indicator data can be explained according to the information entropy, and the corresponding category weight is calculated according to the information entropy. Information entropy is a commonly used method for processing multidimensional data. The larger the information entropy, the more chaotic the system in the group is, and the richer the information contained is. Relatively speaking, this indicator is more meaningful. The calculation expression of the category weight is:
其中，Wj为类别权重，指标的信息熵分别为E1,E2,…,En；对得到的类别权重进行检验，确保其合理性和可靠性，检验完成后便可得到对应的类别权重；基于对应的类别权重建立在预设单位时间中各类危险驾驶行为的单位距离风险，采用面积法，基于对应的类别权重建立单位时间内超过限速值的速度所对应的面积与其对应速度风险之间的联系，将所有行驶时间内的速度风险进行累加求和，即可得到针对超速危险驾驶行为的路段驾驶总风险，并将其除以路段距离得到其对应的单位距离风险，各类危险驾驶行为的单位距离风险都是按上述处理，直至将所有危险驾驶行为的单位距离风险都得到为止；基于所述单位距离风险利用交叉指数法构建交叉指数矩阵，结合单位距离风险通过expm函数构建交叉指数矩阵，expm函数是一种用于计算指数矩阵的函数；基于所述交叉指数矩阵量化各类危险驾驶行为的相对风险；对各类危险驾驶行为对应的类别权重和相对风险进行变量替换，变量替换是为了处理类别权重和相对风险计算中的非线性问题，基于变量替换后的类别权重和相对风险计算驾驶行为得分，通过计算类别权重和危险驾驶行为的相对风险来计算驾驶行为得分，能够更为客观地评定计算驾驶员的驾驶行为得分。Among them,Wj is the category weight, and the information entropy of the indicators isE1 ,E2 , …,En respectively; the obtained category weight is tested to ensure its rationality and reliability, and the corresponding category weight can be obtained after the test is completed; based on the corresponding category weight, the unit distance risk of various dangerous driving behaviors in the preset unit time is established, and the area method is used to establish the relationship between the area corresponding to the speed exceeding the speed limit value in the unit time and its corresponding speed risk based on the corresponding category weight. The speed risks in all driving times are accumulated and summed to obtain the total road section driving risk for speeding dangerous driving behaviors, and it is divided by the road section distance to obtain its corresponding unit distance risk. The unit distance risks of various dangerous driving behaviors are processed as above until the unit distance risks of all dangerous driving behaviors are obtained. To the end; construct a cross-index matrix based on the unit distance risk using the cross-index method, and construct a cross-index matrix through the expm function in combination with the unit distance risk, the expm function is a function used to calculate the index matrix; quantify the relative risks of various types of dangerous driving behaviors based on the cross-index matrix; perform variable replacement on the category weights and relative risks corresponding to various types of dangerous driving behaviors, the variable replacement is to deal with the nonlinear problems in the calculation of category weights and relative risks, calculate the driving behavior score based on the category weights and relative risks after variable replacement, and calculate the driving behavior score by calculating the category weights and the relative risks of dangerous driving behaviors, which can more objectively evaluate and calculate the driver's driving behavior score.
综合风险得分计算模块25：用于获取驾驶员的心理素质数据，基于所述心理素质数据利用预设要素测评值计算驾驶员的心理状况得分，并基于所述心理状况得分和驾驶行为得分计算驾驶员的综合风险得分；Comprehensive risk score calculation module 25: used to obtain the driver's psychological quality data, calculate the driver's psychological condition score based on the psychological quality data using the preset factor evaluation value, and calculate the driver's comprehensive risk score based on the psychological condition score and the driving behavior score;
在本发明具体实施过程中，所述获取驾驶员的心理素质数据，基于所述心理素质数据利用预设要素测评值计算驾驶员的心理状况得分，并基于所述心理状况得分和驾驶行为得分计算驾驶员的综合风险得分，包括：采集驾驶员的心理素质数据，所述心理素质数据包括心理认知能力、人格特质和心理健康状况；对驾驶员的心理认知能力、人格特质和心理健康状况分别赋予对应的加权权重，基于对应的加权权重利用预设要素测评值计算驾驶员的心理状况得分；基于所述心理状况得分和驾驶行为得分利用预设得分比例计算驾驶员的综合风险得分。In the specific implementation process of the present invention, the driver's psychological quality data is obtained, the driver's psychological condition score is calculated based on the psychological quality data using preset factor evaluation values, and the driver's comprehensive risk score is calculated based on the psychological condition score and the driving behavior score, including: collecting the driver's psychological quality data, the psychological quality data includes psychological cognitive ability, personality traits and mental health status; assigning corresponding weighted weights to the driver's psychological cognitive ability, personality traits and mental health status, respectively, and calculating the driver's psychological condition score based on the corresponding weighted weights using preset factor evaluation values; calculating the driver's comprehensive risk score based on the psychological condition score and the driving behavior score using a preset score ratio.
具体的，通过对接对应的管理平台采集驾驶员的心理素质数据，心理素质数据包括心理认知能力、人格特质和心理健康状况；对驾驶员的心理认知能力、人格特质和心理健康状况分别赋予对应的加权权重，根据预设标注你对每项数据都赋予不同的加权权重，通过预设要素测评值对心理认知能力、人格特质和心理健康状况对应的加权权重计算驾驶员的心理状况得分，通过要素测评值能够从不同的维度来评定驾驶员的心理状况得分，基于所述心理状况得分和驾驶行为得分利用预设得分比例计算驾驶员的综合风险得分，心理状况得分和驾驶行为得分分别与对应的预设比例相乘，将两者相乘的结果再相加，得到驾驶员的综合风险得分，将驾驶员的心理状况得分也加入至风险得分的考虑中，通过由心理状况得分和驾驶行为得分所生成的综合风险得分能够更为准确且客观地对驾驶员进行风险评价。Specifically, the psychological quality data of the driver is collected by connecting to the corresponding management platform, and the psychological quality data includes psychological cognitive ability, personality traits and mental health status; corresponding weighted weights are assigned to the driver's psychological cognitive ability, personality traits and mental health status respectively, and different weighted weights are assigned to each data according to the preset annotations, and the driver's psychological condition score is calculated by the preset factor evaluation value and the weighted weights corresponding to the psychological cognitive ability, personality traits and mental health status. The factor evaluation value can be used to evaluate the driver's psychological condition score from different dimensions, and the driver's comprehensive risk score is calculated based on the psychological condition score and the driving behavior score using the preset score ratio, the psychological condition score and the driving behavior score are respectively multiplied by the corresponding preset ratios, and the results of the multiplication of the two are added together to obtain the driver's comprehensive risk score, and the driver's psychological condition score is also added into the risk score consideration, and the comprehensive risk score generated by the psychological condition score and the driving behavior score can more accurately and objectively evaluate the risk of the driver.
管控模块26：用于基于驾驶员的综合风险得分对驾驶员进行风险等级评级，基于驾驶员的风险等级生成对应的管控策略，并将管控策略发送至对应的驾驶员。The control module 26 is used to rate the risk level of the driver based on the driver's comprehensive risk score, generate a corresponding control strategy based on the driver's risk level, and send the control strategy to the corresponding driver.
在本发明具体实施过程中，所述基于驾驶员的综合风险得分对驾驶员进行风险等级评级，基于驾驶员的风险等级生成对应的管控策略，并将管控策略发送至对应的驾驶员，包括：基于驾驶员的综合风险得分利用预设评级标准对驾驶员划分风险等级；基于驾驶员的风险等级利用对驾驶员所标定的各类危险驾驶行为生成对应的管控策略，并将对应的管控策略实时发送至对应的驾驶员中。During the specific implementation of the present invention, the driver's risk level is rated based on the driver's comprehensive risk score, a corresponding management and control strategy is generated based on the driver's risk level, and the management and control strategy is sent to the corresponding driver, including: dividing the driver's risk level based on the driver's comprehensive risk score using preset rating standards; generating corresponding management and control strategies based on the driver's risk level using various types of dangerous driving behaviors calibrated for the driver, and sending the corresponding management and control strategies to the corresponding driver in real time.
具体的，基于驾驶员的综合风险得分利用预设评级标准对驾驶员划分风险等级，根据不同的综合风险得分分别划分为低风险等级、中低风险等级、中风险等级、中高风险等级和高风险；基于驾驶员的风险等级利用对驾驶员所标定的各类危险驾驶行为生成对应的管控策略，对驾驶员做出相应的具体管理限制和行为更正，如驾驶员的风险等级为高风险，则需要对应的驾驶员每隔一个时段便进行汇报，根据其所发生的危险驾驶行为进行更正，直至驾驶员的风险等级降低至预设标准，才解除对驾驶员的相应管理限制。Specifically, based on the comprehensive risk score of the driver, the driver is divided into risk levels using preset rating standards, and is divided into low risk level, medium-low risk level, medium risk level, medium-high risk level and high risk according to different comprehensive risk scores; based on the driver's risk level, a corresponding management and control strategy is generated for each type of dangerous driving behavior calibrated for the driver, and corresponding specific management restrictions and behavioral corrections are made to the driver. If the driver's risk level is high risk, the corresponding driver is required to report at regular intervals and make corrections based on the dangerous driving behavior that has occurred until the driver's risk level is reduced to the preset standard, and then the corresponding management restrictions on the driver are lifted.
在本发明实施例中，根据对驾驶行为数据的特征分析结果利用预设特征阈值标定各类危险行为，将危险驾驶行为进行具体的标定分类，再通过计算类别权重和危险驾驶行为的相对风险来计算驾驶行为得分，能够更为客观地评定计算驾驶员的驾驶行为得分，并将驾驶员的心理状况得分也加入至风险得分的考虑中，通过由心理状况得分和驾驶行为得分所生成的综合风险得分能够更为准确且客观地对驾驶员进行风险评价，从而能够更为准确且具体地选取对应的管控策略，对驾驶员的管理能够达到更好的效果。In an embodiment of the present invention, various types of dangerous behaviors are calibrated using preset characteristic thresholds based on the characteristic analysis results of driving behavior data, and the dangerous driving behaviors are specifically calibrated and classified. Then, the driving behavior score is calculated by calculating the category weight and the relative risk of the dangerous driving behavior. The driving behavior score of the driver can be evaluated and calculated more objectively, and the driver's psychological condition score is also added to the risk score consideration. The comprehensive risk score generated by the psychological condition score and the driving behavior score can more accurately and objectively evaluate the risk of the driver, so that the corresponding control strategy can be selected more accurately and specifically, and the management of the driver can achieve better results.
本领域普通技术人员可以理解上述实施例的各种方法中的全部或部分步骤是可以通过程序来指令相关的硬件来完成，该程序可以存储于一计算机可读存储介质中，存储介质可以包括：只读存储器(ROM，ReadOnly Memory)、随机存取存储器(RAM，Random AccessMemory)、磁盘或光盘等。A person skilled in the art may understand that all or part of the steps in the various methods of the above embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.
另外，以上对本发明实施例所提供的一种驾驶员风险评价管理方法及系统进行了详细介绍，本文中应采用了具体个例对本发明的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本发明的方法及其核心思想；同时，对于本领域的一般技术人员，依据本发明的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本发明的限制。In addition, the above describes in detail a driver risk assessment management method and system provided by an embodiment of the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410066155.9ACN118025193A (en) | 2024-01-17 | 2024-01-17 | Driver risk evaluation management method and system |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410066155.9ACN118025193A (en) | 2024-01-17 | 2024-01-17 | Driver risk evaluation management method and system |
| Publication Number | Publication Date |
|---|---|
| CN118025193Atrue CN118025193A (en) | 2024-05-14 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410066155.9APendingCN118025193A (en) | 2024-01-17 | 2024-01-17 | Driver risk evaluation management method and system |
| Country | Link |
|---|---|
| CN (1) | CN118025193A (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118243133A (en)* | 2024-05-28 | 2024-06-25 | 名商科技有限公司 | Driving path optimized navigation method and system based on road condition environment recognition |
| CN120171540A (en)* | 2025-05-21 | 2025-06-20 | 江西江铃集团新能源汽车有限公司 | Driver driving style determination method, device, electronic device and storage medium |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140278574A1 (en)* | 2013-03-14 | 2014-09-18 | Ernest W. BARBER | System and method for developing a driver safety rating |
| CN110648075A (en)* | 2019-09-27 | 2020-01-03 | 重庆大学 | Driving safety assessment method and device |
| CN110866677A (en)* | 2019-10-25 | 2020-03-06 | 东南大学 | A method for evaluating the relative risk of drivers based on benchmarking analysis |
| CN113689071A (en)* | 2021-07-15 | 2021-11-23 | 东南大学 | An active intervention method for bad driving behavior based on multi-driver risk assessment |
| CN116611621A (en)* | 2023-07-18 | 2023-08-18 | 枣庄卡企安网络科技有限公司 | Multi-role real-time data interaction and supervision traffic safety management system |
| CN116738312A (en)* | 2023-06-08 | 2023-09-12 | 山东大学 | Dangerous driving behavior identification and early warning method and system based on vehicle driving data |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140278574A1 (en)* | 2013-03-14 | 2014-09-18 | Ernest W. BARBER | System and method for developing a driver safety rating |
| CN110648075A (en)* | 2019-09-27 | 2020-01-03 | 重庆大学 | Driving safety assessment method and device |
| CN110866677A (en)* | 2019-10-25 | 2020-03-06 | 东南大学 | A method for evaluating the relative risk of drivers based on benchmarking analysis |
| CN113689071A (en)* | 2021-07-15 | 2021-11-23 | 东南大学 | An active intervention method for bad driving behavior based on multi-driver risk assessment |
| CN116738312A (en)* | 2023-06-08 | 2023-09-12 | 山东大学 | Dangerous driving behavior identification and early warning method and system based on vehicle driving data |
| CN116611621A (en)* | 2023-07-18 | 2023-08-18 | 枣庄卡企安网络科技有限公司 | Multi-role real-time data interaction and supervision traffic safety management system |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118243133A (en)* | 2024-05-28 | 2024-06-25 | 名商科技有限公司 | Driving path optimized navigation method and system based on road condition environment recognition |
| CN120171540A (en)* | 2025-05-21 | 2025-06-20 | 江西江铃集团新能源汽车有限公司 | Driver driving style determination method, device, electronic device and storage medium |
| Publication | Publication Date | Title |
|---|---|---|
| CN118025193A (en) | Driver risk evaluation management method and system | |
| CN110703214B (en) | Weather radar state evaluation and fault monitoring method | |
| CN102663264B (en) | Semi-supervised synergistic evaluation method for static parameter of health monitoring of bridge structure | |
| CN110619482A (en) | Driving behavior scoring method based on logistic regression and single-level analysis weighting method | |
| CN111275338A (en) | Method, device, equipment and storage medium for judging enterprise fraud behaviors | |
| CN111222968A (en) | Enterprise tax risk management and control method and system | |
| CN119003994B (en) | Clinical trial data risk early warning system based on big data analysis | |
| CN107348964B (en) | Psychological load measurement method of drivers in extra-long tunnel environment based on factor analysis | |
| CN116957361B (en) | Ship task system health state detection method based on virtual-real combination | |
| CN109784586B (en) | Prediction method and system for danger emergence condition of vehicle danger | |
| CN113505980A (en) | Reliability evaluation method, device and system for intelligent traffic management system | |
| CN116701130A (en) | Dynamic baseline optimization method and device based on index portrait and electronic equipment | |
| CN119417646A (en) | Secondary water supply system fault early warning method and system based on data analysis | |
| CN111341106B (en) | Traffic early warning method, device and equipment | |
| CN118861293B (en) | A method for identifying dangerous driving behaviors of drivers based on text data | |
| CN109978396A (en) | A kind of early screening system and method for risk case | |
| CN118942606B (en) | Integrated intelligent integrated management and control system for hospital information | |
| CN117952482B (en) | Product quality accident grading method and system based on convolutional neural network | |
| CN114444872A (en) | Method and system for analyzing driving behavior of driver by adopting artificial intelligence algorithm | |
| CN118296533A (en) | A bridge influence line identification method based on empirical variational mixed mode decomposition | |
| CN115437350B (en) | Signal system health evaluation method and device based on fault analysis table | |
| CN114613130B (en) | Driving credibility analysis method in traffic and carrying system | |
| CN113962530B (en) | An indicator early warning method and system based on data mining | |
| CN113807587B (en) | Integral early warning method and system based on multi-ladder nuclear deep neural network model | |
| CN116739617A (en) | Food related product risk management system and method based on data analysis |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication | Application publication date:20240514 |