CN110689643A - An Immune Algorithm-Based Vehicle Driving State Analysis Method for Intelligent Networked Vehicles - Google Patents
An Immune Algorithm-Based Vehicle Driving State Analysis Method for Intelligent Networked VehiclesDownload PDFInfo
- Publication number
- CN110689643A CN110689643ACN201910906549.XACN201910906549ACN110689643ACN 110689643 ACN110689643 ACN 110689643ACN 201910906549 ACN201910906549 ACN 201910906549ACN 110689643 ACN110689643 ACN 110689643A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- state
- driving
- antigen
- driving state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C5/00—Registering or indicating the working of vehicles
- G07C5/08—Registering or indicating performance data other than driving, working, idle, or waiting time, with or without registering driving, working, idle or waiting time
- G07C5/0808—Diagnosing performance data
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/002—Biomolecular computers, i.e. using biomolecules, proteins, cells
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C5/00—Registering or indicating the working of vehicles
- G07C5/008—Registering or indicating the working of vehicles communicating information to a remotely located station
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C5/00—Registering or indicating the working of vehicles
- G07C5/08—Registering or indicating performance data other than driving, working, idle, or waiting time, with or without registering driving, working, idle or waiting time
- G07C5/0816—Indicating performance data, e.g. occurrence of a malfunction
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Artificial Intelligence (AREA)
- Organic Chemistry (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Chemical & Material Sciences (AREA)
- Traffic Control Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Translated fromChineseDescription
Translated fromChinese技术领域technical field
本发明属于智能网联汽车道路交通主动安全技术领域，具体涉及一种基于免 疫算法的智能网联汽车车辆行驶状态分析方法。The invention belongs to the technical field of road traffic active safety of intelligent networked vehicles, in particular to a method for analyzing the driving state of intelligent networked vehicles based on an immune algorithm.
背景技术Background technique
自动驾驶发展到现在，自动驾驶汽车车辆行驶状态的监测，自动驾驶汽车行 驶的安全性一直是大家关心的问题，智能网联汽车，通过车载摄像头、激光雷达、 毫米波雷达、卫星定位等技术，实施智能算法实现行驶车辆的环境感知，通过车 载的总线系统以及车上的其他各种功能的传感器可以获得当前车辆在行驶过程 中的一些状态信息参数，而智能网联汽车V2X技术，使得汽车和汽车、汽车和 人、汽车和道路、汽车和远程监控平台能够实时的进行通讯，因此，这些技术使 得智能车辆上的所有信息可以方便的传送到相关监控平台，并将监控车辆的行驶 状态及时的传达给相关的车、人和道路等。Since the development of autonomous driving, the monitoring of the driving status of autonomous vehicles and the safety of autonomous vehicles have always been a concern for everyone. Intelligent networked vehicles, through in-vehicle cameras, lidar, millimeter-wave radar, satellite positioning and other technologies, Implement intelligent algorithms to realize the environment perception of the driving vehicle. Some state information parameters of the current vehicle during the driving process can be obtained through the on-board bus system and other sensors of various functions on the vehicle. Cars, cars and people, cars and roads, cars and remote monitoring platforms can communicate in real time. Therefore, these technologies enable all information on smart vehicles to be easily transmitted to relevant monitoring platforms, and monitor the driving status of vehicles in a timely manner. Communicate to relevant cars, people and roads.
发明内容SUMMARY OF THE INVENTION
本发明所要解决的技术问题在于针对上述现有技术中的不足，提供一种基于 免疫算法的智能网联汽车车辆行驶状态分析方法，以智能汽车视觉、雷达、定位 的环境感知系统为基础，通过该系统和车载总线及其他传感器，实时获取智能汽 车行驶过程中的有关数据，并传送到云端服务器，在云服务器端以免疫算法建立 模型，对行驶车辆的行驶状态进行评价，当发现有不安全的状态时，及时给与预 警，提高了道路交通安全主动防控能力。The technical problem to be solved by the present invention is to provide a method for analyzing the driving state of an intelligent networked vehicle based on an immune algorithm, which is based on the environment perception system of intelligent vehicle vision, radar and positioning, The system, the on-board bus and other sensors, acquire the relevant data in real time during the driving process of the smart car, and transmit it to the cloud server. On the cloud server, an immune algorithm is used to build a model to evaluate the driving state of the driving vehicle. When the state of road traffic safety is in a state of emergency, early warning is given in time, which improves the ability of active prevention and control of road traffic safety.
本发明采用以下技术方案：The present invention adopts following technical scheme:
一种基于免疫算法的智能网联汽车车辆行驶状态分析方法，包括以下步骤：A method for analyzing the driving state of an intelligent networked vehicle based on an immune algorithm, comprising the following steps:
S1、以汽车行驶状态评价知识库作为自体库，生成每种车辆状态对应的检测 器，以自体库作为已知抗原，激活对应的B细胞，由B细胞产生与抗原相对应 的抗体，当未知抗原产生时，激活相关抗体，根据激活抗体的概率推导出车辆当 前所处于的状态；S1. Use the vehicle driving state evaluation knowledge base as the autologous library, generate detectors corresponding to each vehicle state, use the autologous library as the known antigen, activate the corresponding B cells, and the B cells produce antibodies corresponding to the antigens. When the antigen is generated, the relevant antibody is activated, and the current state of the vehicle is deduced according to the probability of activating the antibody;
S2、输入非自体库和检测器距离，计算小于阈值的距离，确定每种典型状态 的概率，对车辆行驶状态进行评价。S2. Input the distance between the non-self library and the detector, calculate the distance less than the threshold, determine the probability of each typical state, and evaluate the driving state of the vehicle.
具体的，步骤S1具体为：Specifically, step S1 is specifically:
S101、抗原识别，生产自体库；S101, antigen recognition, production of autologous library;
S102、生成B细胞，进行亲和力计算；S102, generate B cells, and perform affinity calculation;
S103、记忆库更新、抗体的抑制和促进，如果大于设定阈值，返回步骤S102 重新计算亲和力，如果小于阈值，生成检测器。S103, memory bank update, antibody inhibition and promotion, if it is greater than the set threshold, return to step S102 to recalculate the affinity, and if it is less than the threshold, generate a detector.
进一步的，步骤S101中，自体集由典型的车辆行驶状态的数据组成，作为 已知抗原，一组车辆行驶状态的数据建立一个一维向量集合，第i种车辆行驶状 态，j为车辆状态参数个数为：Further, in step S101, the self-set is composed of data of typical vehicle driving states. As a known antigen, a set of vehicle driving state data establishes a one-dimensional vector set, the i-th vehicle driving state, and j is the vehicle state parameter. The number is:
code[i]＝{ci1,ci2,ci3……cij}code[i]={ci1,ci2,ci3...cij}
设c1：车速，c2：加速度，c3：电池温度，c4：车与前方车辆的距离，c5： 车道偏离次数，车道偏离次数是每分钟车辆行驶偏离车道的次数Let c1: vehicle speed, c2: acceleration, c3: battery temperature, c4: distance between the car and the vehicle ahead, c5: the number of lane departures, the number of lane departures is the number of times the vehicle deviates from the lane per minute
code[i]＝{ci1,ci2,ci3,ci4,ci5}code[i]={ci1,ci2,ci3,ci4,ci5}
而由n组车辆行驶状态数据构成一个n维自体库向量集合，i＝1～n，j＝1～5；On the other hand, an n-dimensional self-library vector set is formed by n groups of vehicle driving state data, i=1～n, j=1～5;
非自体输入为车辆在行驶过程中的任一状态，作为未知抗原：Non-autologous input is any state of the vehicle during driving, as an unknown antigen:
uncode[i]＝{ui1,ui2……uij}uncode[i]={ui1,ui2...uij}
和自体库一样：As with the autologous library:
uncode[i]＝{ui1,ui2,ui3,ui4,ui5}uncode[i]＝{ui1,ui2,ui3,ui4,ui5}
而由n组车辆行驶状态数据构成一个n维非自体库向量集合：i＝1～n，j＝1～5However, an n-dimensional non-self library vector set is formed by n groups of vehicle driving state data: i=1～n, j=1～5
进一步的，步骤S102中，根据建立的典型的车辆行驶状态参数作为抗原， 设某一个状态参数为向量code[i]＝[ci1,ci2,…cij]，以该抗原为中心生成一个新的B 细胞，Bi(x1,x2…xj)；对于B细胞，生成数量为N的抗体，设每个个体大小为 k个，则抗体数量为：Further, in step S102, according to the established typical vehicle driving state parameters as an antigen, a certain state parameter is set as a vector code[i]=[ci1,ci2,...cij], and a new B is generated with the antigen as the center Cells, Bi(x1,x2...xj); for B cells, the number of antibodies produced is N, and the size of each individual is set to be k, then the number of antibodies is:
N＝j*kN=j*k
初始抗体的产生来源有两种，如问题在记忆库中有所保留，则取记忆库，不 足部分随机生成，若记忆库为空则全部随机生成，当抗原入侵机体时，B细胞被 激活以识别特异抗原，此时B细胞大量繁殖：There are two sources of initial antibody generation. If the problem remains in the memory bank, the memory bank will be taken, and the insufficient part will be randomly generated. If the memory bank is empty, all the antibodies will be randomly generated. When the antigen invades the body, the B cells are activated to Recognizing specific antigens, B cells multiply at this time:
B[i]＝{bi1,bi2,…bij}B[i]={bi1,bi2,...bij}
经过学习生成新的B细胞为：After learning, new B cells are generated as:
newB[i]＝{nbi1,nbi2…nbij}newB[i]={nbi1,nbi2...nbij}
采用Euclidean距离计算亲和力克隆因子：The affinity cloning factor is calculated using Euclidean distance:
进一步的，步骤S103中，生成的每种车辆状态对应的检测器为：Further, in step S103, the generated detectors corresponding to each vehicle state are:
其中，i代表第i个典型车辆行驶状态，k是种群的大小。Among them, i represents the ith typical vehicle driving state, and k is the size of the population.
具体的，步骤S2中，当监测器生成后，以车辆行驶过程的任一状态作为非 自体输入，计算出该未知抗原与n个监测器的Euclidean距离D，设定一个距离 的阈值m，计算出每个距离di中dij<m的距离的累加和M，得出该抗体属于该 监测器的概率Pi，根据概率的最大值判断出该未知抗原所属的已知抗原的类别， 对车辆的行驶状态作出评价。Specifically, in step S2, after the monitor is generated, the Euclidean distance D between the unknown antigen and the n monitors is calculated by using any state of the vehicle driving process as the non-autologous input, and a distance threshold m is set to calculate Calculate the cumulative sum M of the distances of dij<m in each distance di, obtain the probability Pi that the antibody belongs to the monitor, and determine the category of the known antigen to which the unknown antigen belongs according to the maximum value of the probability. status evaluation.
进一步的，监测器的概率Pi为：Further, the probability Pi of the monitor is:
Pi＝M/di1+di2+…+dij。Pi=M/di1+di2+...+dij.
进一步的，定义评价包括安全、较安全、不安全和危险，未知抗原与n个监 测器的Euclidean距离D为：Further, the definition evaluation includes safe, relatively safe, unsafe and dangerous, and the Euclidean distance D between the unknown antigen and n monitors is:
D＝{d1,d2…di}D={d1,d2...di}
其中，i＝1～n，n＝4，di为未知抗原参数与第i个监测器的距离。Among them, i=1∼n, n=4, and di is the distance between the unknown antigen parameter and the ith monitor.
与现有技术相比，本发明至少具有以下有益效果：Compared with the prior art, the present invention at least has the following beneficial effects:
本发明一种基于免疫算法的智能网联汽车车辆行驶状态分析方法，以自主研 发的智能网联汽车车辆信息采集系统为平台，实时获取智能汽车当前行驶过程中 有关的数据，并传送到云端服务器，在云端服务器以智能免疫遗传算法建立模型， 对智能汽车当前的行驶状态进行评判。本发明所采用的免疫遗传算法是将免疫理 论和基本遗传算法各自的优点结合起来的一个多学科相互交叉、渗透的优化算 法，并将其应用于智能网联汽车车辆行驶安全状态的分析，由于该免疫遗传算法 既保留了免疫算法的优点，又提高了免疫算法中抗体的多样性和收敛速度。最终 通过实验对模型做了验证，对数据进行了分析，结果表明了这种模型对智能网联 汽车行驶安全状态评价的可行性。The present invention is a method for analyzing the driving state of an intelligent networked vehicle based on an immune algorithm. The self-developed intelligent networked vehicle vehicle information collection system is used as a platform to obtain real-time data related to the current driving process of the intelligent vehicle and transmit it to a cloud server. , build a model on the cloud server with an intelligent immune genetic algorithm to judge the current driving state of the smart car. The immune genetic algorithm adopted in the present invention is a multi-disciplinary and interpenetrating optimization algorithm that combines the respective advantages of the immune theory and the basic genetic algorithm, and is applied to the analysis of the driving safety state of the intelligent networked vehicle. The immune genetic algorithm not only retains the advantages of the immune algorithm, but also improves the diversity and convergence speed of antibodies in the immune algorithm. Finally, the model is verified by experiments, and the data is analyzed. The results show the feasibility of this model to evaluate the driving safety state of intelligent networked vehicles.
进一步的，以典型的车辆行驶状态建立自体库，产生已知抗原，在自体库已 知抗原的基础上生成B细胞，随机生成的方式产生抗体种群，以抗体与已知抗原 的距离计算抗体的亲和力，并以此作为适应度，通过遗传算法进化种群，包括交 叉、变异、选择，通过亲和力控制迭代过程，最后生成针对每种车辆典型行驶状 态抗原的监测器。通过生成的监测器以及车辆实时行驶的状态参数作为未知抗 原，就可以对当前车辆行驶状态进行评判。Further, an autologous library is established in a typical vehicle driving state, known antigens are generated, B cells are generated on the basis of the known antigens in the autologous library, and antibody populations are generated randomly, and the distance between the antibody and the known antigen is calculated. The affinity is used as the fitness, and the population is evolved through genetic algorithm, including crossover, mutation, selection, and the iterative process is controlled by affinity, and finally the monitor for each vehicle's typical driving state antigen is generated. By using the generated monitor and the real-time driving state parameters of the vehicle as unknown antigens, the current vehicle driving state can be judged.
进一步的，以典型的车辆行驶状态的数据作为已知抗原，一组车辆行驶状态 的数据建立一个一维向量集合，第i种车辆行驶状态，j为车辆状态参数个数为： code[i]＝{ci1,ci2,ci3……cij}；设c1：车速，c2：加速度，c3：电池温度，c4：车 与前方车辆的距离，c5：车道偏离次数，车道偏离次数是每分钟车辆行驶偏离车 道的次数：code[i]＝{ci1,ci2,ci3,ci4,ci5}；则由n组车辆行驶状态数据构成一个n 维自体库向量集合，i＝1～n，j＝1～5；Further, taking the data of the typical vehicle driving state as the known antigen, a set of vehicle driving state data establishes a one-dimensional vector set, the i-th vehicle driving state, j is the number of vehicle state parameters: code[i] ={ci1,ci2,ci3...cij}; set c1: vehicle speed, c2: acceleration, c3: battery temperature, c4: distance between the vehicle and the vehicle ahead, c5: lane departure times, the lane departure times is the vehicle's departure per minute. Number of lanes: code[i]={ci1,ci2,ci3,ci4,ci5}; then an n-dimensional self-library vector set is formed by n groups of vehicle driving state data, i=1～n, j=1～5;
自体库是典型的汽车行驶状态评价知识库，它作为系统的已知抗原，才能激 活B细胞学习并产生与抗原相对应的抗体。The autologous library is a typical vehicle driving state evaluation knowledge base. As the known antigen of the system, it can activate B cells to learn and produce antibodies corresponding to the antigen.
进一步的，根据建立的典型的车辆行驶状态参数作为抗原，设某一个状态参 数为向量code[i]＝[ci1,ci2,…cij]，以该抗原为中心生成一个新的B细胞，Bi(x1,x2… xj)；对于B细胞，生成数量为N的抗体，设每个个体大小为k个，则抗体数量 为：N＝j*k；当抗原入侵机体时，B细胞被激活以识别特异抗原，此时B细胞大 量繁殖：B[i]＝{bi1,bi2,…bij}；经过学习生成新的B细胞为：newB[i]＝{nbi1,nbi2… nbij}；采用Euclidean距离计算亲和力克隆因子：Further, according to the established typical vehicle driving state parameters as the antigen, set a certain state parameter as the vector code[i]=[ci1,ci2,...cij], and generate a new B cell with the antigen as the center, Bi( x1,x2...xj); for B cells, the number of antibodies generated is N, and the size of each individual is set to be k, then the number of antibodies is: N=j*k; when the antigen invades the body, B cells are activated to recognize Specific antigen, B cells multiply at this time: B[i]={bi1,bi2,…bij}; after learning to generate new B cells: newB[i]={nbi1,nbi2…nbij}; use Euclidean distance calculation Affinity cloning factor:
B细胞的生成是为了产生与抗原相对应的抗体，然后以抗体与已知抗原的距 离计算抗体的亲和力，亲和力代表抗体和抗原的匹配原则，即识别强度。留下亲 和度高的抗体进行存储，以保证种群向着适应度好的方向进化。The generation of B cells is to produce antibodies corresponding to the antigen, and then the affinity of the antibody is calculated by the distance between the antibody and the known antigen, and the affinity represents the matching principle between the antibody and the antigen, that is, the recognition strength. Keep antibodies with high affinity for storage to ensure that the population evolves in the direction of good fitness.
进一步的，以抗体与已知抗原的距离计算出的亲和力作为适应度，通过遗传 算法进化种群，包括交叉、变异、选择，通过设定的阈值控制迭代过程，最后生 成针对每种车辆典型行驶状态抗原的监测器。应用该监测器就可以对车辆行驶过 程中任一行驶状态进行评判。Further, the affinity calculated from the distance between the antibody and the known antigen is used as the fitness, and the population is evolved through the genetic algorithm, including crossover, mutation, selection, and the iterative process is controlled by the set threshold, and finally the typical driving state of each vehicle is generated. Antigen monitor. The application of the monitor can evaluate any driving state during the driving process of the vehicle.
进一步的，根据生成的监测器，以车辆行驶过程的任一状态作为非自体输入， 计算出该未知抗原与n个监测器的Euclidean距离D，设定一个距离的阈值m， 计算出每个距离di中dij<m的距离的累加和M，得出该抗体属于该监测器的概 率Pi，根据概率的最大值判断出该未知抗原所属的已知抗原的类别，对车辆的行 驶状态作出评价。Further, according to the generated monitor, use any state of the vehicle driving process as the non-self input, calculate the Euclidean distance D between the unknown antigen and the n monitors, set a distance threshold m, and calculate each distance. The cumulative sum M of the distances of dij<m in di, the probability Pi that the antibody belongs to the monitor is obtained, and the class of the known antigen to which the unknown antigen belongs is judged according to the maximum probability, and the driving state of the vehicle is evaluated.
综上所述，本发明以自主研发的智能网联汽车车辆信息采集系统为平台，实 时获取智能汽车当前行驶过程中状态参数，并传送到云端服务器，在云端服务器 以智能免疫算法建立模型，对智能汽车当前的行驶状态进行评判，当发现有不安 全的状态时，及时给与预警，最后通过实验对模型做了验证，对数据进行了分析， 结果表明了这种模型对智能网联汽车行驶状态评价的可行性。本发明适用于各种 交通状况以及各种车型，提高了汽车在行驶过程中的主动安全性能，降低了事故 发生率。并且驾驶员可以通过车载终端实时了解车辆的相关状态参数，便于对车 辆进行评估和保养；管理人员也可以通过控制中心掌握车辆运行状况，加强车辆 管理，做出高效率的调度。To sum up, the present invention takes the self-developed intelligent network-connected vehicle vehicle information collection system as a platform, acquires the state parameters of the current driving process of the intelligent vehicle in real time, and transmits it to the cloud server. The current driving state of the smart car is judged, and an early warning is given in time when an unsafe state is found. Finally, the model is verified through experiments, and the data is analyzed. Feasibility of condition evaluation. The invention is suitable for various traffic conditions and various vehicle types, improves the active safety performance of the vehicle during the driving process, and reduces the accident rate. In addition, the driver can understand the relevant state parameters of the vehicle in real time through the on-board terminal, which is convenient for the evaluation and maintenance of the vehicle; the management personnel can also grasp the operating status of the vehicle through the control center, strengthen the vehicle management, and make efficient scheduling.
下面通过附图和实施例，对本发明的技术方案做进一步的详细描述。The technical solutions of the present invention will be further described in detail below through the accompanying drawings and embodiments.
附图说明Description of drawings
图1为车辆行驶状态免疫评价模型图；Fig. 1 is the model diagram of immune evaluation of vehicle driving state;
图2为B细胞与抗体学习过程图；Figure 2 is a diagram of the learning process of B cells and antibodies;
图3为自体库生成检测器的流程图；Fig. 3 is the flow chart of autologous library generation detector;
图4为非自体库输入，车辆行驶状态评价流程图；Fig. 4 is the non-self library input, the flow chart of vehicle driving state evaluation;
图5为未知抗原非自体的输入图；Fig. 5 is the input diagram of unknown antigen non-autologous;
图6为群体总适应度平均值变化图；Figure 6 is a graph showing the average change of the total fitness of the population;
图7为评判结果图。Figure 7 is a graph of the evaluation results.
具体实施方式Detailed ways
请参阅图1和图2，生物免疫系统是一个庞大而复杂的系统，它涉及很多细 胞分子，根据行驶车辆安全性的特点，设计了一种车辆行驶状态免疫评价模型， 通过采集车辆上的数据并传送到云端服务器，对正在行驶的车辆状态进行监测。Please refer to Figure 1 and Figure 2. The biological immune system is a huge and complex system, which involves many cells and molecules. According to the safety characteristics of the driving vehicle, an immune evaluation model for the driving state of the vehicle is designed. By collecting the data on the vehicle And send it to the cloud server to monitor the status of the running vehicle.
自体库为典型的汽车行驶状态评价知识库，作为系统的已知抗原，由建立的 典型的安全性评价数据自体库作为已知抗原，激活对应的B细胞，从而由B细 胞产生与抗原相对应的抗体，当未知抗原产生时，激活所相关的抗体，根据激活 抗体的概率推导出车辆当前所处于的状态。The autologous library is a typical vehicle driving state evaluation knowledge base. As the known antigen of the system, the established typical safety evaluation data autologous library is used as the known antigen to activate the corresponding B cells, so that the B cells produce corresponding antigens. When the unknown antigen is produced, the relevant antibody is activated, and the current state of the vehicle is deduced according to the probability of activating the antibody.
智能网联汽车车辆行驶状态诊断免疫模型的状态空间由反映车辆行驶的几 个主要信息参数构成，可以描述出抗体和抗原之间的相互作用，系统状态可以用 特征向量表示，对于求解过程中出现的各种状态分别建立相应的状态空间。The state space of the ICV vehicle driving state diagnosis immune model is composed of several main information parameters reflecting the driving of the vehicle, which can describe the interaction between antibodies and antigens, and the system state can be represented by eigenvectors. The various states of , respectively, establish corresponding state spaces.
请参阅图3和图4，本发明一种基于免疫算法的智能网联汽车车辆行驶状态 分析方法，包括以下步骤：Referring to Fig. 3 and Fig. 4 , a method for analyzing the driving state of an intelligent networked vehicle based on an immune algorithm of the present invention includes the following steps:
S1、通过自体库生成检测器；S1. Generate a detector through an autologous library;
S101、抗原识别，生产自体库；S101, antigen recognition, production of autologous library;
初始化自体库Initialize self library
输入典型的车辆的状态参数作为自体输入，初始化记忆库：Enter the state parameters of a typical vehicle as self input to initialize the memory bank:
code＝[i][j]code=[i][j]
其中，i为某个典型车辆状态，j为该状态对应的车辆参数。Among them, i is a typical vehicle state, and j is the vehicle parameter corresponding to this state.
因状态和参数较多，现挑选几个举例说明，设车辆状态评价分为为“安全、 较安全、不安全、危险”，对应的参数选择“车速、加速度、温度、与前方车辆 的距离、车道偏离次数”，以车辆行驶过程的任一状态参数uncode[i]作为非自体 输入，对其行驶状态进行评价。Due to the large number of states and parameters, here are a few examples to illustrate. Let the vehicle state evaluation be divided into “safe, relatively safe, unsafe, dangerous”, and the corresponding parameters can be selected as “vehicle speed, acceleration, temperature, distance from the vehicle ahead, The number of lane departures”, taking any state parameter uncode[i] of the vehicle’s driving process as a non-self input, to evaluate its driving state.
自体集与非自体输入Self-set and non-self input
自体集由典型的车辆行驶状态的数据组成，作为已知抗原，一组车辆行驶状 态的数据建立一个一维向量集合，下面是第i种车辆行驶状态，j为车辆状态参数 个数。The self-set is composed of typical vehicle driving state data. As a known antigen, a set of vehicle driving state data establishes a one-dimensional vector set. The following is the i-th vehicle driving state, and j is the number of vehicle state parameters.
code[i]＝{ci1,ci2,ci3……cij}code[i]={ci1,ci2,ci3...cij}
因为采集到的车辆状态参数很多，由于篇幅的限制，以其中几个参数为例来 说明算法的原理，设Because there are many vehicle state parameters collected, due to space limitations, several parameters are used as examples to illustrate the principle of the algorithm.
c1：车速，c2：加速度，c3：电池温度，c4：车与前方车辆的距离，c5：车 道偏离次数，车道偏离次数是每分钟车辆行驶偏离车道的次数。c1: vehicle speed, c2: acceleration, c3: battery temperature, c4: distance between the vehicle and the vehicle ahead, c5: lane departure times, the lane departure times is the number of times the vehicle deviates from the lane per minute.
按照举例的情况则：As an example:
code[i]＝{ci1,ci2,ci3,ci4,ci5}code[i]={ci1,ci2,ci3,ci4,ci5}
而由n组车辆行驶状态数据构成一个n维自体库向量集合，i＝1～n，j＝1～5；On the other hand, an n-dimensional self-library vector set is formed by n groups of vehicle driving state data, i=1～n, j=1～5;
非自体输入为车辆在行驶过程中的任一状态，作为未知抗原：Non-autologous input is any state of the vehicle during driving, as an unknown antigen:
uncode[i]＝{ui1,ui2……uij}uncode[i]={ui1,ui2...uij}
和自体库一样，以其中几个参数为例来说明算法的原理，则：Like the autologous library, taking several parameters as an example to illustrate the principle of the algorithm, then:
uncode[i]＝{ui1,ui2,ui3,ui4,ui5}uncode[i]＝{ui1,ui2,ui3,ui4,ui5}
而由n组车辆行驶状态数据构成一个n维非自体库向量集合：i＝1～n，j＝1～5However, an n-dimensional non-self library vector set is formed by n groups of vehicle driving state data: i=1～n, j=1～5
S102、生成B细胞，进行亲和力计算；S102, generate B cells, and perform affinity calculation;
定义B细胞及初始抗体的生成Defining B cells and initial antibody production
根据之前建立的典型的车辆行驶状态参数作为抗原，设某一个状态参数为向 量code[i]＝[ci1,ci2,…cij]，以该抗原为中心生成一个新的B细胞，Bi(x1,x2…xj)。According to the typical vehicle driving state parameters established before as the antigen, set a certain state parameter as the vector code[i]=[ci1,ci2,...cij], and generate a new B cell with the antigen as the center, Bi(x1, x2...xj).
对于B细胞，生成数量为N的抗体，设每个个体大小为k个，则抗体数量 为：For B cells, the number of antibodies produced is N, and if the size of each individual is k, the number of antibodies is:
N＝j*kN=j*k
本发明中举例参数个数为5个，个体为20个，则抗体数为100个。In the present invention, the number of parameters is 5, the number of individuals is 20, and the number of antibodies is 100.
初始抗体的产生来源有两种，如问题在记忆库中有所保留，则取记忆库，不 足部分随机生成，若记忆库为空则全部随机生成。There are two sources for the generation of initial antibodies. If the problem remains in the memory bank, the memory bank will be taken, and the insufficient part will be randomly generated. If the memory bank is empty, all of the antibodies will be randomly generated.
当抗原入侵机体时，B细胞被激活以识别特异抗原，此时B细胞大量繁殖。When the antigen invades the body, B cells are activated to recognize the specific antigen, and the B cells multiply at this time.
B[i]＝{bi1,bi2,…bij}B[i]={bi1,bi2,...bij}
经过学习生成新的B细胞为：After learning, new B cells are generated as:
newB[i]＝{nbi1,nbi2…nbij}newB[i]={nbi1,nbi2...nbij}
采用Euclidean距离计算亲和力克隆因子Calculation of affinity cloning factor using Euclidean distance
亲和力代表抗体和抗原的匹配原则，即识别强度，根据抗体与所属B细胞中 心的距离来计算亲和力，一般亲和力的计算公式为：Affinity represents the matching principle between antibody and antigen, that is, the recognition strength. The affinity is calculated according to the distance between the antibody and the center of the B cell to which it belongs. The general affinity calculation formula is:
式中，tk为抗原与抗体的结合强度，一般免疫算法计算结合强度tk的数学工 具主要有：In the formula, tk is the binding strength of antigen and antibody, and the mathematical tools for calculating binding strength tk of general immune algorithms mainly include:
海明距离：Hamming distance:
Euclidean距离：Euclidean distance:
Manhattan距离：Manhattan Distance:
S103、记忆库更新、抗体的抑制和促进，如果大于设定阈值，返回步骤S102 重新计算亲和力，如果小于阈值，生成检测器。S103, memory bank update, antibody inhibition and promotion, if it is greater than the set threshold, return to step S102 to recalculate the affinity, and if it is less than the threshold, generate a detector.
选择亲和度高的抗体进行存储，更新记忆库，因而，亲和度高的抗体显然受 到促进，传进下一代的概率更大，而亲和度低的就会受到抑制。Antibodies with high affinity are selected for storage and the memory bank is updated. Therefore, antibodies with high affinity are obviously promoted and have a greater probability of being passed on to the next generation, while those with low affinity will be inhibited.
抗体Antibody
每一个B细胞都被设定(基因编码)产生一种特质的抗体，抗体是B细胞 识别抗原后增殖分化为浆细胞所产生的，一个B细胞只产生一种特异抗体。Each B cell is programmed (genetically encoded) to produce a specific antibody. The antibody is produced by the proliferation and differentiation of the B cell into plasma cells after recognizing the antigen. A B cell produces only one specific antibody.
学习过程中整个种群去掉适应度低的，留下适应度高的，每次迭代后生成的 第i种车辆状态的分类器为During the learning process, the whole population will remove the ones with low fitness, and leave the ones with high fitness. The classifier of the i-th vehicle state generated after each iteration is
其中，k为种群大小，j为车辆状态参数个数，设每个基因二进制编码位数为 c-lemgth。Among them, k is the population size, j is the number of vehicle state parameters, and the number of bits in the binary code of each gene is set as c-lemgth.
遗传操作genetic manipulation
由于亲和度高的抗体受到促进，亲和度低的就会受到抑制，这样很容易导致 群体进化单一，导致局部优化，因此需要在算法中引入新的策略，保证群体的多 样性，本发明的遗传操作是应用交叉、变异产生下一代抗体，保证种族的多样性， 并向着适应度好的方向进化。Since antibodies with high affinity are promoted, those with low affinity are inhibited, which can easily lead to a single population evolution and local optimization. Therefore, it is necessary to introduce a new strategy into the algorithm to ensure the diversity of the population. The present invention The genetic manipulation is to use crossover and mutation to generate the next generation of antibodies, to ensure the diversity of races, and to evolve in the direction of good fitness.
检测器的生成Generation of detectors
经过交叉变异后的种群，按照适应度进行排序，将低适应度的个体去掉，留 下高适应度的，更新记忆库，如此反复迭代下去，直到群体中最大适应度的个体， 本发明中适应度是进化中种群抗体与自体库抗原的距离相关值，距离越近适应度 越小，所以当适应度小于设定的阈值，则可以停止进化，则该车辆状态的监测器 生成，如此类推，有关车辆的n组状态的监测器就可生成。After cross-mutation, the population is sorted according to fitness, the individuals with low fitness are removed, and those with high fitness are left, and the memory bank is updated, and so on repeatedly until the individual with the largest fitness in the group is adapted in the present invention. The degree is the distance correlation value between the population antibody and the autologous library antigen in the evolution. The shorter the distance, the smaller the fitness. Therefore, when the fitness is less than the set threshold, the evolution can be stopped, and the monitor of the vehicle status will be generated, and so on. Monitors for n sets of states of the vehicle can then be generated.
当迭代完成后，生成的每种车辆状态对应的检测器为：When the iteration is completed, the generated detectors corresponding to each vehicle state are:
其中，i代表第i个典型车辆行驶状态，k是种群的大小。Among them, i represents the ith typical vehicle driving state, and k is the size of the population.
S2、输入非自体库，对车辆行驶状态进行评价S2. Input the non-self library to evaluate the driving state of the vehicle
S201、输入非自体和检测器距离；S201. Input the distance between the non-self and the detector;
S202、计算小于阈值的距离；S202, calculating a distance less than a threshold;
S203、确定每种典型状态的概率，完成车辆行驶状态评价。S203: Determine the probability of each typical state, and complete the vehicle running state evaluation.
对车辆行驶过程的任一状态的评判：Judgment on any state of the vehicle driving process:
当监测器生成后，就可应用监测器对任意的车辆行驶状态进行评判。After the monitor is generated, the monitor can be used to judge any vehicle driving state.
以车辆行驶过程的任一状态作为非自体输入，计算出该未知抗原与n个监测 器的Euclidean距离D：Taking any state of the vehicle driving process as the non-autologous input, calculate the Euclidean distance D between the unknown antigen and n monitors:
D＝{d1,d2…di}i＝1～nD={d1,d2...di}i=1～n
本发明中共有“安全、较安全、不安全、危险”4个评价，则n＝4，di为未 知抗原参数与第i个监测器的距离。In the present invention, there are 4 evaluations of "safe, relatively safe, unsafe, dangerous", then n=4, and di is the distance between the unknown antigen parameter and the i-th monitor.
因个体数为k个,设j＝1～k，则：Since the number of individuals is k, set j=1～k, then:
d1＝{d11,d12,…d1j}d1={d11,d12,…d1j}
d2＝{d21,d22,…d2j}d2={d21,d22,...d2j}
： :
di＝{di1,di2,….dij}di={di1,di2,….dij}
本发明产生个体个数k为20个。The present invention produces 20 individuals, k.
设定一个距离的阈值m，计算出每个距离di中dij<m的距离的累加和M， 因而就得出该抗体属于该监测器的概率：Set a distance threshold m, calculate the cumulative sum M of the distances of dij<m in each distance di, and thus obtain the probability that the antibody belongs to the monitor:
Pi＝M/di1+di2+…+dijPi=M/di1+di2+…+dij
根据概率的最大值判断出该未知抗原所属的已知抗原的类别，从而对车辆的 该行驶状态作出评价。The class of the known antigen to which the unknown antigen belongs is determined according to the maximum value of the probability, so as to evaluate the driving state of the vehicle.
本发明研发了智能汽车基于视觉、雷达、定位的环境感知系统，通过该系统 和车载总线以及其他传感器，实时获取智能汽车行驶过程中的有关数据，并传送 到云端服务器，在云服务器端以免疫算法建立模型，对行驶车辆的行驶状态进行 评价，当发现有不安全的状态时，及时给与预警。The invention develops an environment perception system based on vision, radar and positioning for smart cars. Through the system, on-board bus and other sensors, relevant data during the driving process of smart cars are acquired in real time, and sent to the cloud server. The algorithm establishes a model, evaluates the driving state of the driving vehicle, and gives an early warning in time when an unsafe state is found.
为使本发明实施例的目的、技术方案和优点更加清楚，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 描述的实施例是本发明一部分实施例，而不是全部的实施例。通常在此处附图中 的描述和所示的本发明实施例的组件可以通过各种不同的配置来布置和设计。因 此，以下对在附图中提供的本发明的实施例的详细描述并非旨在限制要求保护的 本发明的范围，而是仅仅表示本发明的选定实施例。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他实施例，都属 于本发明保护的范围。In order to make the purposes, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments These are some embodiments of the present invention, but not all embodiments. The components of the embodiments of the invention generally described and illustrated in the drawings herein may be arranged and designed in a variety of different configurations. Thus, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the invention as claimed, but rather to represent only selected embodiments of the invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present invention.
本发明一种基于免疫算法的智能网联汽车车辆行驶状态分析方法，实验分别 在长安大学渭水校区智能网联汽车实验场和西安绕城高速上进行，实验设备及现 场如图5、图6所示。The present invention is a method for analyzing the driving state of an intelligent networked vehicle based on an immune algorithm. The experiments are carried out on the intelligent networked vehicle experimental site of Chang'an University Weishui Campus and the Xi'an Ring Expressway. The experimental equipment and site are shown in Figures 5 and 6. Show.
实验设备包括图像采集及处理系统，激光雷达测距与环境感知系统，GPS定 位系统，另外数据来源还包括车上的CAN线以及车上的其他传感器，所有的数据 都直接或通过车载终端发送到云端服务器进行处理和存储。The experimental equipment includes an image acquisition and processing system, a lidar ranging and environmental perception system, and a GPS positioning system. In addition, the data source also includes the CAN line on the vehicle and other sensors on the vehicle. All data are sent directly or through the vehicle terminal. Cloud server for processing and storage.
设已知自体库如下：Let the known self library be as follows:
表1.自体库(已知抗原)Table 1. Autologous repertoire (known antigens)
4组数据分别代表车辆行驶状态为“安全、较安全、不安全、危险”。The 4 sets of data respectively represent the driving status of the vehicle as "safe, relatively safe, unsafe, dangerous".
以实验所得数据作为未知抗原非自体的输入，从实验数据中列取50组来说 明，如图7所示。Taking the data obtained from the experiment as the input of the unknown antigen non-autologous, 50 groups were selected from the experimental data to illustrate, as shown in Figure 7.
设个体种群个数k为20，基因二进制编码位数c-lemgth为4，初始抗体种群随机生成， 经过选择、交叉、变异，迭代次数100次，群体总适应度平均值变化如图6所示。Let the number of individual populations k be 20, the number of bits of gene binary coding c-lemgth be 4, the initial antibody population is randomly generated, after selection, crossover, mutation, and the number of iterations is 100, the average change of the total fitness of the population is shown in Figure 6. .
以种群与自体库的距离作为适应度，从图中看出，随着迭代次数的增加适应 度不断减小，解达到收敛。Taking the distance between the population and the self-storage as the fitness, it can be seen from the figure that as the number of iterations increases, the fitness decreases and the solution converges.
以生成的监测器对50组数据(图7中的行驶状态)计算出每组数据针对自体库中已知抗 原的概率，最大概率对应的状态就是这组数据评判结果，得到的评判结果如图7。Using the generated monitor to calculate the probability of each group of data for known antigens in the autologous library for 50 sets of data (driving state in Figure 7), the state corresponding to the maximum probability is the evaluation result of this group of data, and the obtained evaluation result is shown in the figure 7.
从图中列举10组典型数据来说明评判结果，输入未知抗原如表2：10 groups of typical data are listed in the figure to illustrate the judgment results, and the unknown antigens are input as shown in Table 2:
表2输入未知抗原Table 2 Input unknown antigen
对表2的10组数据监测器输出的评判结果见表3，The judging results of the output of the 10 groups of data monitors in Table 2 are shown in Table 3,
表3监测器输出Table 3 Monitor outputs
表中数据为未知抗原属于已知抗原的概率，从表中可以看出，数据1车速低、 距离较大，是安全的的行驶状态，数据6因为距离较小，是较安全状态，数据4 出现了两个相邻状态一样的概率，这种情况结论就定位处于安全与较安全之间 (所以图3中有状态处于1.5和2.5的情况)，数据5因为加速度较大，偏离次 数较多，是不安全状态，数据9因为车速较高，偏离次数多，所以是危险的状态。The data in the table is the probability that the unknown antigen belongs to the known antigen. It can be seen from the table that the
输出数据中还出现了非相邻数据出现了相同最高概率的情况，对于这种情况 评判为安全性低的状态，并将数据记录为待进一步验证数据。In the output data, there is also a situation where the non-adjacent data has the same highest probability. For this situation, it is judged as a low security state, and the data is recorded as the data to be further verified.
本发明针对智能网联汽车车辆行驶状态的特点，设计了基于免疫算法的车辆 行驶状态监测系统，通过我们自己研发的设备进行实验获取数据，以典型的数据 作为自体库，以任意行驶状态数据作为非自体库，对车辆行驶状态进行评价，经 过验证免疫算法对车辆行驶状态的评价是基本正确的，结论的可靠性还与自体库 的完整性、初始种群的选择、迭代次数以及阈值的选择有很大的相关性，另外实 验获取数据的稳定性对输出数据也有影响，经过更多的验证和改进算法并与其他 的智能算法相结合，可以更大的增加结论的可靠性。Aiming at the characteristics of the driving state of intelligent networked vehicles, the present invention designs a vehicle driving state monitoring system based on an immune algorithm, and obtains data through experiments with equipment developed by ourselves. The non-autologous database is used to evaluate the driving state of the vehicle. It has been verified that the evaluation of the driving state of the vehicle by the immune algorithm is basically correct. The reliability of the conclusion is also related to the integrity of the autologous database, the selection of the initial population, the number of iterations and the selection of the threshold. In addition, the stability of the experimental data has an impact on the output data. After more verification and improvement of the algorithm and combination with other intelligent algorithms, the reliability of the conclusion can be greatly increased.
以上内容仅为说明本发明的技术思想，不能以此限定本发明的保护范围，凡 是按照本发明提出的技术思想，在技术方案基础上所做的任何改动，均落入本发 明权利要求书的保护范围之内。The above content is only to illustrate the technical idea of the present invention, and cannot limit the protection scope of the present invention. Any changes made on the basis of the technical solution according to the technical idea proposed by the present invention all fall within the scope of the claims of the present invention. within the scope of protection.
Claims (8)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910906549.XACN110689643B (en) | 2019-09-24 | 2019-09-24 | Intelligent networking automobile vehicle driving state analysis method based on immune algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910906549.XACN110689643B (en) | 2019-09-24 | 2019-09-24 | Intelligent networking automobile vehicle driving state analysis method based on immune algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110689643Atrue CN110689643A (en) | 2020-01-14 |
| CN110689643B CN110689643B (en) | 2022-07-26 |
Family
ID=69109961
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910906549.XAExpired - Fee RelatedCN110689643B (en) | 2019-09-24 | 2019-09-24 | Intelligent networking automobile vehicle driving state analysis method based on immune algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110689643B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112581654A (en)* | 2020-12-29 | 2021-03-30 | 华人运通(江苏)技术有限公司 | System and method for evaluating use frequency of vehicle functions |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663167A (en)* | 2012-03-20 | 2012-09-12 | 浙江大学 | Optimization design method for electric automobile anti-lock braking system controller based on immune algorithm |
| CN102867409A (en)* | 2012-09-17 | 2013-01-09 | 李伟 | Road traffic cooperative control method for urban central area |
| CN203480000U (en)* | 2013-10-15 | 2014-03-12 | 浙江大学城市学院 | Detector for health status of power lithium battery for full electric vehicle |
| CN204405692U (en)* | 2014-12-30 | 2015-06-17 | 上海研亚软件信息技术有限公司 | Judge the device of vehicle movement status |
| CN105954607A (en)* | 2015-11-12 | 2016-09-21 | 北京交通大学 | Method and system for detecting faults of high-speed railway signal system |
| CN107784587A (en)* | 2016-08-25 | 2018-03-09 | 大连楼兰科技股份有限公司 | A Driving Behavior Evaluation System |
| CN108572324A (en)* | 2018-04-13 | 2018-09-25 | 芜湖职业技术学院 | Battery SOC Estimation Device Based on Immune Algorithm Optimizing BP Neural Network |
| CN108961473A (en)* | 2018-08-07 | 2018-12-07 | 长安大学 | A kind of vehicle-state assessment method for early warning based on intelligent network connection automobile control centre |
| US20190197414A1 (en)* | 2017-12-22 | 2019-06-27 | At&T Intellectual Property I, L.P. | System and method for estimating potential injuries from a vehicular incident |
| EP3505051A1 (en)* | 2017-12-29 | 2019-07-03 | Sanmina Corporation | Vehicular health monitoring system and method |
| DE102018203606A1 (en)* | 2018-03-09 | 2019-09-12 | Bayerische Motoren Werke Aktiengesellschaft | High-voltage energy storage device |
- 2019
- 2019-09-24CNCN201910906549.XApatent/CN110689643B/ennot_activeExpired - Fee Related
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663167A (en)* | 2012-03-20 | 2012-09-12 | 浙江大学 | Optimization design method for electric automobile anti-lock braking system controller based on immune algorithm |
| CN102867409A (en)* | 2012-09-17 | 2013-01-09 | 李伟 | Road traffic cooperative control method for urban central area |
| CN203480000U (en)* | 2013-10-15 | 2014-03-12 | 浙江大学城市学院 | Detector for health status of power lithium battery for full electric vehicle |
| CN204405692U (en)* | 2014-12-30 | 2015-06-17 | 上海研亚软件信息技术有限公司 | Judge the device of vehicle movement status |
| CN105954607A (en)* | 2015-11-12 | 2016-09-21 | 北京交通大学 | Method and system for detecting faults of high-speed railway signal system |
| CN107784587A (en)* | 2016-08-25 | 2018-03-09 | 大连楼兰科技股份有限公司 | A Driving Behavior Evaluation System |
| US20190197414A1 (en)* | 2017-12-22 | 2019-06-27 | At&T Intellectual Property I, L.P. | System and method for estimating potential injuries from a vehicular incident |
| EP3505051A1 (en)* | 2017-12-29 | 2019-07-03 | Sanmina Corporation | Vehicular health monitoring system and method |
| DE102018203606A1 (en)* | 2018-03-09 | 2019-09-12 | Bayerische Motoren Werke Aktiengesellschaft | High-voltage energy storage device |
| CN108572324A (en)* | 2018-04-13 | 2018-09-25 | 芜湖职业技术学院 | Battery SOC Estimation Device Based on Immune Algorithm Optimizing BP Neural Network |
| CN108961473A (en)* | 2018-08-07 | 2018-12-07 | 长安大学 | A kind of vehicle-state assessment method for early warning based on intelligent network connection automobile control centre |
Non-Patent Citations (7)
| Title |
|---|
| LIGUOWENG: "Immune network-based swarm intelligence and its application to unmanned aerial vehicle (UAV) swarm coordination", 《NEUROCOMPUTING》* |
| LIGUOWENG: "Immune network-based swarm intelligence and its application to unmanned aerial vehicle (UAV) swarm coordination", 《NEUROCOMPUTING》, 11 February 2014 (2014-02-11), pages 134 - 141* |
| 刘洋: "基于免疫机制的智能车路径规划与运动控制研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》* |
| 刘洋: "基于免疫机制的智能车路径规划与运动控制研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》, 15 September 2014 (2014-09-15), pages 035 - 65* |
| 孟庆华等: "疫苗算法及其在车辆故障检测中的应用", 《浙江大学学报(工学版)》, no. 02, 28 February 2006 (2006-02-28)* |
| 滕凯凯,仝秋红,吕奎超: "基于BP神经网络的车道偏离预警系统的阈值与照度关系的研究", 《汽车实用技术》* |
| 滕凯凯,仝秋红,吕奎超: "基于BP神经网络的车道偏离预警系统的阈值与照度关系的研究", 《汽车实用技术》, 30 April 2016 (2016-04-30), pages 36 - 38* |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112581654A (en)* | 2020-12-29 | 2021-03-30 | 华人运通(江苏)技术有限公司 | System and method for evaluating use frequency of vehicle functions |
| CN112581654B (en)* | 2020-12-29 | 2022-09-30 | 华人运通(江苏)技术有限公司 | System and method for evaluating use frequency of vehicle functions |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110689643B (en) | 2022-07-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113954864B (en) | Intelligent automobile track prediction system and method integrating peripheral automobile interaction information | |
| CN111930110B (en) | An intention trajectory prediction method combined with social generative adversarial network | |
| US11673550B2 (en) | Agent prioritization for autonomous vehicles | |
| Althoff et al. | Set-based prediction of traffic participants on arbitrary road networks | |
| CN110928297B (en) | Path planning method for intelligent public transport vehicles based on multi-objective dynamic particle swarm optimization | |
| CN112258841B (en) | Intelligent vehicle risk assessment method based on vehicle track prediction | |
| CN107272687A (en) | A kind of driving behavior decision system of automatic Pilot public transit vehicle | |
| Liu et al. | An inverse reinforcement learning approach for customizing automated lane change systems | |
| Zhang et al. | An ensemble learning–online semi-supervised approach for vehicle behavior recognition | |
| CN114644018B (en) | A game theory-based decision-making and planning method for human-vehicle interaction in autonomous driving vehicles | |
| Li et al. | Development and evaluation of two learning-based personalized driver models for pure pursuit path-tracking behaviors | |
| CN116118780A (en) | Vehicle obstacle avoidance track planning method, system, vehicle and storage medium | |
| Ma et al. | Driving style estimation by fusing multiple driving behaviors: a case study of freeway in China | |
| Prezioso et al. | Machine learning insights for behavioral data analysis supporting the autonomous vehicles scenario | |
| CN110689643B (en) | Intelligent networking automobile vehicle driving state analysis method based on immune algorithm | |
| Naitmalek et al. | Embedded real-time speed forecasting for electric vehicles: a case study on RSK urban roads | |
| CN116523158A (en) | Multi-unmanned aerial vehicle track planning method, device, equipment and storage medium | |
| Hu et al. | A safe driving decision-making methodology based on cascade imitation learning network for automated commercial vehicles | |
| Cheng et al. | Towards Safe Motion Planning for Autonomous Driving in Highway | |
| CN115092138A (en) | A lane-changing trajectory planning method for vehicles on expressways based on natural driver characteristics | |
| Jiang et al. | EDRF: Enhanced Driving Risk Field Based on Multimodal Trajectory Prediction and its Applications | |
| Wang et al. | A data-driven estimation of driving style using deep clustering | |
| Su et al. | Learning vehicle cooperative lane-changing behavior from observed trajectories in the NGSIM dataset | |
| Li et al. | Multi-Step Continuous Decision Making and Planning in Uncertain Dynamic Scenarios Through Parallel Spatio-Temporal Trajectory Searching | |
| Huang et al. | Digital twin edge services with proximity-aware longitudinal lane changing model for connected vehicles |
Legal Events
| 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | Granted publication date:20220726 | |
| CF01 | Termination of patent right due to non-payment of annual fee |