Detailed Description
By providing the intelligent detection early warning method of the traffic overload system, the technical problems that overload behaviors cannot be timely, comprehensively and accurately identified due to low detection efficiency, limited coverage range and lack of effective monitoring on dynamic changes of vehicle loading states of a fixed wagon balance detection station in the prior art are solved, and the efficiency and safety of road traffic management are further affected. The technical targets of dynamic, real-time and efficient monitoring and anomaly identification of the load state of the vehicle are realized, and the technical effects of comprehensively improving the road traffic management efficiency and reducing the influence of overload behaviors on road infrastructure and traffic safety are achieved.
In the following, the technical solutions of the present application will be clearly and completely described with reference to the accompanying drawings, and it should be understood that the described embodiments are only some embodiments of the present application, but not all embodiments of the present application, and that the present application is not limited by the exemplary embodiments described herein. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application. It should be further noted that, for convenience of description, only some, but not all of the drawings related to the present application are shown.
Referring to fig. 1, the application provides an intelligent detection and early warning method of a traffic control system, which specifically comprises the following steps of firstly, arranging a sensor group in a pre-test area, collecting data of a vehicle, and establishing a vehicle data set.
Specifically, the pre-test area refers to an area where the sensor device is defined and set in advance before traffic is overtime, so that all the monitored information is more accurate. A sensor group is a device composed of a plurality of sensors for collecting various data related to a vehicle, such as a vehicle speed, a license plate number, a loading state, and the like. And arranging a sensor group in the pre-test area, and collecting data of the vehicle by using the sensor group so as to effectively monitor the vehicle entering the preset area.
Next, the collected data is established as a vehicle dataset. The vehicle data set is a database storing all information related to a particular vehicle, such as visual information including the speed of travel of the vehicle in the pre-test area, license plate data, and vehicle loading. The vehicle running speed, namely vehicle speed data, is acquired through a speed sensor in a sensor group, such as a radar or a laser velocimeter, license plate data is acquired through a license plate recognition technology, such as a license plate recognition camera in the sensor group, and visual information of vehicle loading, namely vehicle loading images, is acquired through an image acquisition device in the sensor group, such as a camera or an image sensor. And establishing a vehicle data set, further evaluating whether the vehicle is overloaded or not through subsequent analysis and processing, and further performing intelligent early warning.
And secondly, activating a load sensor in the test area, and configuring the test frequency of the load sensor through the vehicle speed data in the vehicle data set.
Specifically, the test area refers to an actual monitoring area in the process of traffic overload. The load sensor is a sensor that measures the weight of the vehicle. In the test area, the load sensor is activated to bring the load sensor into operation for measuring the vehicle weight.
The test frequency of the load sensor is then configured with the vehicle speed data in the vehicle data set. The configuration test frequency refers to adjusting the sampling frequency of the load sensor according to the vehicle speed so as to ensure that the load sensor can accurately acquire the vehicle weight data. If the vehicle speed is faster, the load sensor needs a higher sampling frequency to capture accurate weight data, otherwise, the load sensor is lower, so that the data accuracy is ensured, and the energy waste caused by excessive work of the load sensor is avoided.
And thirdly, establishing a load data set by using the configured load sensor, wherein the load data set is provided with a time mark.
Specifically, after the load sensor is subjected to early debugging and setting, such as sampling frequency, weight information of the vehicle is collected and stored, and a load data set is established. For example, when the load sensor is installed on a road surface or a bridge and a vehicle passes through, the load sensor senses the weight of the vehicle and converts the weight into data, so that the weight acquisition is completed.
The load data set refers to weight information related data acquired by the load sensor, and comprises information such as weight, passing time and the like of the vehicle, namely the load data set is provided with a time mark. The time stamp includes the specific time of data acquisition to ensure timeliness and accuracy of the data.
And step four, acquiring speed measurement data in the test area based on the time mark, wherein the speed measurement data comprises real-time speed data and acceleration data.
Specifically, according to the time identifier recorded in the load data set, vehicle speed information at the corresponding moment of the time identifier is extracted from the speed sensors in the sensor group in the test area, and speed measurement data are obtained. The speed measurement data refers to the related information of the running speed of the vehicle, and can reflect the dynamic behavior of the vehicle in the test area. The speed measurement data includes real-time speed data and acceleration data. The real-time speed data is the running speed of the vehicle at a specific time, and is usually continuously collected in units of time, so that the current speed state of the vehicle can be reflected. Acceleration data represents the rate of change of speed over time and is used to describe the dynamics of acceleration or deceleration of a vehicle. Acceleration is the derivative of the velocity change and can be obtained by calculating the difference between adjacent velocity data.
And fifthly, inputting the speed measurement data and the load data set into a load compensation model, and establishing a load measurement result.
Specifically, the speed measurement data and the load data set are input into a load compensation model, and are calculated and processed through the load compensation model, so that a more accurate load measurement result is finally obtained. The load compensation model is used for correcting or compensating load data according to environmental parameters (such as temperature, humidity and the like). The load compensation model can adjust load data according to the influence of environmental factors and vehicle dynamics. And outputting final load measurement data after processing the load compensation model, and establishing a load measurement result.
And step six, calling vehicle information according to license plate data, establishing a first early warning according to a vehicle information calling result and the load measuring result, carrying out image recognition on a vehicle loading image, carrying out linkage analysis according to an image recognition result and the load measuring result, establishing a second early warning, and carrying out intelligent detection abnormality report according to the first early warning and the second early warning.
Specifically, vehicle information calling is carried out according to license plate data, unique identification information of a vehicle is obtained through a license plate recognition technology, and then the detailed information of the vehicle is searched from a database by utilizing the unique identification information of the vehicle, so that the vehicle information is obtained. The vehicle information includes vehicle type, manufacturer, owner, history of the vehicle, etc. for use in subsequent detection and analysis.
And establishing a first early warning by using the calling result of the vehicle information and the load measurement result, and combining the vehicle information with the load measurement result after acquiring the vehicle information so as to judge whether overload or other loading conditions which do not meet the regulations exist. The load measurement result is vehicle load data measured by a sensor, an instrument, or the like. If the load of the vehicle exceeds the specified standard, an early warning is established and relevant personnel are reminded.
The method comprises the steps of shooting a loading image of a vehicle through equipment such as a camera, analyzing the loading image of the vehicle by utilizing an image recognition technology, and recognizing articles loaded on the vehicle and distribution conditions of the articles. The image recognition technology generally classifies and measures cargoes in images, checks whether the cargoes accord with loading standards, and assists in judging whether loading of vehicles is problematic through visual recognition.
And carrying out linkage analysis according to the image recognition result and the load measurement result, and after obtaining the image recognition result and the load measurement data, combining the information of the image recognition result and the load measurement data to carry out comprehensive analysis, so as to establish a second early warning. For example, image recognition may indicate that a vehicle is loaded with a large volume of cargo, and load measurements may indicate that the load of the vehicle has approached or exceeded a standard, thereby more accurately determining whether an overload or other loading problem exists through linkage analysis. The second warning means that if a loading abnormality or load discrepancy is found, a higher level warning is generated. For example, if the first warning only shows a load deviation, the second warning may indicate that there is a severe overload or unstable loading, thereby facilitating a more urgent process. And carrying out intelligent detection abnormality report according to the first early warning and the second early warning, further confirming whether abnormality exists or not through an intelligent detection algorithm, and automatically reporting the abnormality.
The intelligent detection early warning method of the traffic control system can realize the technical targets of dynamic, real-time and efficient monitoring and anomaly identification of the load state of the vehicle, and achieve the technical effects of comprehensively improving the road traffic management efficiency and reducing the influence of overload behaviors on road infrastructure and traffic safety.
Further referring to FIG. 2, the method further comprises synchronizing the load data set to an environment compensation layer of the load compensation model, performing environment monitoring of the test area by using an environment monitoring sensor, establishing an environment data set, wherein the environment data set comprises temperature data, humidity data and road surface state data, synchronizing the environment data set to the environment compensation layer, executing load data set compensation, and synchronizing the compensated load data set and the speed measurement data to a calculation layer to generate a load measurement result.
Specifically, the load compensation model is used for processing errors related to the load, so that more accurate weight data is ensured. The environment compensation layer is a layer in the load compensation model and is used for integrating the influence of environment factors such as temperature, humidity and the like on load data and ensuring the data precision. Wherein a large amount of basic data is collected, including raw load data of the vehicle, environmental data (such as temperature, humidity, road surface state, etc.), and other relevant parameters. The basic data are collected in real time through equipment such as a load sensor, an environment monitoring sensor and the like which are arranged in a test area. A proper machine learning or statistical modeling method is selected to train a load-carrying compensation model, such as linear regression, support Vector Machines (SVM), neural networks and the like, and the error of the load-carrying sensor is predicted and compensated according to input basic data. After training, partial basic data divided in advance is used for verifying the load compensation model, and whether the compensation effect of the load compensation model reaches the expected value is evaluated. And synchronizing the load data set to an environment compensation layer of the load compensation model, namely transmitting the collected vehicle weight data to the environment compensation layer of the load compensation model to carry out load error compensation, and obtaining the compensated load data.
An environmental monitoring sensor is a device in a sensor group for detecting and registering ambient environmental conditions. And (3) performing environmental monitoring on the temperature, the humidity, the road surface condition and the like of the test area by using an environmental monitoring sensor, and establishing an environmental data set, wherein the environmental data set comprises temperature data, humidity data and road surface condition data. The temperature data records the ambient temperature, the humidity data records the moisture content in the air, and the road surface state data reflects the actual conditions of the road, such as whether the road is slippery, whether water is accumulated, and the like.
And synchronizing the environmental data set to an environmental compensation layer, correcting or compensating the load data set by using the environmental data set in the environmental compensation layer, eliminating the interference of environmental factors on the measured data, and improving the accuracy of the data. And synchronously transmitting the compensated load data set and the speed measurement data to a calculation layer to generate a load measurement result. The load measurement result refers to an actual load value of the vehicle obtained after factors such as environmental compensation, speed change and the like are considered, and the actual load value is used for judging whether the vehicle is overloaded or not, or providing basis for other management decisions.
Further, the application also comprises the steps of establishing the parameter dynamic weight as follows: Wherein, the method comprises the steps of,Characterization of the first embodimentThe dynamic weights of the individual environmental parameters,Is the firstSensitivity coefficients for the individual environmental parameters,Is the firstThe real-time values of the individual environmental parameters,And carrying out dynamic weight compensation and interaction compensation based on the parameter dynamic weight, wherein the dynamic weight compensation and interaction compensation are as follows: Wherein, the method comprises the steps of,The dynamic weight compensation is characterized by the fact that,,The basic compensation coefficient is characterized in that,Characterization of the first embodimentAn influence factor function of the individual environmental parameters,The interactive compensation is characterized in that,,Characterization of the first embodimentPersonal environmental parameters and the firstThe interaction coefficient of the individual environmental parameters,Characterization of the first embodimentDynamic weighting of individual environmental parameters.
In particular, the parameter dynamic weight refers to a dynamic weight used in the environmental compensation layer to quantify the extent to which each environmental data set affects the payload data. By setting the parameter dynamic weight, the environment compensation layer can be more flexible, and the environment data set can be adjusted in real time so as to adapt to the change of the environment.
Wherein,Characterization of the first embodimentDynamic weights of individual environmental parameters are used to represent the weights of particular environmental parameters by variables.Is the firstSensitivity coefficients for each environmental parameter are used to quantify the degree of response of each environmental parameter to payload data.Is the firstReal-time values of the respective environmental parameters, the real-time values representing the firstActual values of the individual environmental parameters at the current instant.For the total number of environmental parameters, the total number of environmental parameters refers to the total number of all relevant environmental factors. Wherein,Is an integer greater than 0.
Dynamic weight compensation and interaction compensation are carried out based on the parameter dynamic weight, and a compensated result is obtained. Dynamic weight compensation refers to adjusting load data according to real-time values and sensitivity coefficients of different environment parameters, so that the load data is more accurate. The interaction compensation refers to further consideration of interaction between environment parameters, which means that different environment parameters may interact to jointly influence load data.
Representing dynamic weight compensation, representing the compensation degree of load data, wherein the compensation degree is subjected to basic compensation coefficientAnd the influence of environmental parameters. Basic compensation coefficientIs an overall estimate of the impact of all environmental parameters in the environmental compensation layer, e.g. base compensation coefficientsIs a constant derived from historical data or a priori knowledge.Characterization of the first embodimentAn influence factor function of the individual environmental parameters, representing the firstThe degree of influence of individual environmental parameters on the payload data under different conditions,And may vary with time, temperature, humidity, etc.
Characterization of the first embodimentDynamic weighting of individual environmental parameters.Characterization of interaction Compensation, reflecting the firstThe influence of each environmental parameter on the final compensation result is further represented, and the compensation effect of the interaction influence between the environmental parameters on the load data is further represented by dynamic weightAnd participating in calculation. Wherein,、Is greater than 0 and less than or equal toIs an integer of (a).
Further, the application also comprises the following steps of: Wherein, the method comprises the steps of,Representing the number of data samples that are to be processed,Characterization of the first embodimentParameters in individual samplesIs used as a reference to the value of (a),Characterization of the first embodimentParameters in individual samplesIs used as a reference to the value of (a),、Characterizing parameters respectivelySum parametersIs used for the average value of (a),、Respectively as parametersSum parametersStandard deviation of (2).
In particular, the interaction impact coefficient represents a joint impact between the environmental parameters. And calculating the interaction influence coefficient, namely measuring interaction values among different environment parameters, and acquiring a compensation result of the load data of the synergistic influence of a plurality of environment factors.
Wherein,Representing the number of data samples, for calculating the total number of data points or samples of the interaction coefficient.Characterization of the first embodimentParameters in individual samplesIs used as a reference to the value of (a),Characterization of the first embodimentParameters in individual samplesIs representative of a value for each sample comprising a set of environmental parameters. Wherein,Is an integer greater than 0 and is selected from the group consisting of,Is greater than 0 and less than or equal toIs an integer of (a).
Next, parametersMean value of (1)Parameters (parameters)Mean value of (1)The average value is obtained by adding the parameter values in all samples and dividing by the number of samples. Parameters (parameters)Standard deviation of (2)Parameters (parameters)Standard deviation of (2)Standard deviation of、Respectively as parametersSum parametersThe degree of fluctuation between all data samples and their mean. The standard deviation is an index reflecting the degree of dispersion of data, and the larger the standard deviation is, the wider the range of variation of data is, and the smaller the standard deviation is, the smaller the fluctuation of data is. Through the mean value and the standard deviation, the interaction influence coefficient can consider the distribution characteristics among the environment parameters, and the influence of the common change of the environment parameters on the load data is measured.
The vehicle load deviation early warning method comprises the steps of calling vehicle type information of a vehicle based on license plate data, generating rated load data of the vehicle according to the vehicle type information, comparing the rated load data with load measurement results to establish load deviation values, calling historical load data of the vehicle by using the license plate data to establish vehicle load behavior abnormality coefficients, focusing and identifying the load deviation values by using the vehicle load behavior abnormality coefficients, and establishing the first early warning.
Specifically, vehicle type information of a vehicle is called based on license plate data, and the vehicle type information corresponding to the license plate in a database is queried. The vehicle type information includes detailed information of a manufacturer, model, body type, etc. of the vehicle for determining a rated load of the vehicle. The rated load of the vehicle is the maximum load capacity specified by the vehicle model in design, and represents the maximum load limit value of the vehicle.
And comparing the load with the actual load measured by using the rated load data and the load measurement result, and comparing the actual load measured by the vehicle with the rated load to obtain a load deviation value which represents the difference between the actual load of the vehicle and the designed maximum load bearing capacity of the vehicle.
And calling historical load data of the vehicle by using license plate data, inquiring and acquiring past load records of the vehicle by using the license plate number, and analyzing load behaviors of the vehicle in different time periods to identify the change of load modes of the vehicle. By analyzing the historical load data, the normal fluctuation range of the load behavior of the vehicle is calculated, the degree of deviation from the normal range is evaluated, and the abnormal coefficient of the load behavior of the vehicle is established.
And carrying out attention recognition on the load deviation value by utilizing the abnormal coefficient of the load behavior of the vehicle, namely judging whether the abnormal behavior exists in the vehicle by analyzing the historical load behavior of the vehicle and combining the current load deviation value. If the load deviation value is large, and the historical load behavior abnormality coefficient of the vehicle also indicates that the vehicle is generally stable in load, and the deviation is more prominent, an alarm is sent to indicate that the vehicle may have overload or other non-compliant load behaviors. Further monitoring measures are taken according to preset rules, and if the vehicle is required to stop, detailed inspection is carried out. If the actual load value of the vehicle is close to the rated load upper limit of the vehicle model or has been very close to the design load limit. At this time, although the load deviation value may not be particularly large, since the actual load of the vehicle approaches its load-bearing limit, there is still a possibility that adverse effects may be exerted on the safety of the vehicle, the road safety, and the transport quality. Thus, by making a more stringent response when the load is determined to be approaching a limit, the vehicle is required to stop to receive more careful inspection.
The method comprises the steps of establishing a standard vehicle type image according to vehicle type information, calling loading goods information of a vehicle by utilizing license plate data after vehicle owner authorization is obtained, carrying out loading abnormality identification by utilizing the standard vehicle type image, an image identification result and the loading goods information, establishing loading standard abnormality, establishing a loading volume prediction result based on the standard vehicle type image and the image identification result, carrying out loading prediction according to the loading volume prediction result and the loading goods information, establishing linkage prediction abnormality, and establishing second early warning according to the loading standard abnormality and the linkage prediction abnormality.
Specifically, a standard vehicle model image is created from the vehicle model information, and a standard image model is generated to represent the appearance characteristics and the dimensional specification of the type of vehicle by querying and using the vehicle model data of the vehicle. The vehicle type information generally includes the model number, manufacturer, body type, and the like of the vehicle, and a standardized vehicle image is generated based on the vehicle type information as a reference for subsequent processing and recognition.
After the authorization of the vehicle owner is obtained, the information of the loaded cargos of the vehicle is called by utilizing license plate data on the premise of the agreement of the vehicle owner, and the current or historical information of the loaded cargos of the vehicle is inquired and obtained, wherein the information comprises the type, the weight, the volume and the like of the cargos. Owner authorization is to ensure privacy and compliance, and loading cargo information is important data reflecting the actual load status of the vehicle.
And comparing the standard vehicle model image with the actual vehicle image, and analyzing whether the loading condition of the vehicle meets the specification or not by combining an image recognition technology. The image recognition result is a vehicle image captured by a camera or other visual device. The loading goods information provides detailed data of the loading of the vehicle, and the loading goods information is combined with the standard vehicle model image, the image recognition result and the loading goods information to judge whether the loading abnormality exists in the vehicle, and the loading behavior of the vehicle is recognized whether to accord with the preset loading specification or not, if the loading situation of the vehicle exceeds the specified limit or is not matched with the standard vehicle model image, the loading specification abnormality exists, and the loading specification abnormality is established. Such as overload, uneven loading, or inconsistent cargo types.
Image recognition is carried out through an image recognition technology, an image recognition result is obtained, the volume of the cargoes loaded on the vehicle is calculated through the image recognition result, the volume is compared with the space specification of a standard vehicle type image, a loading volume prediction result is established, and whether the cargo loading space of the vehicle reaches the maximum loading standard or not and whether the cargo distribution is reasonable or not are evaluated.
And carrying out load prediction according to the load volume prediction result and the load information, wherein the load prediction is used for predicting whether the vehicle is overloaded or whether the potential overload risk exists. When the deviation between the actual load of the vehicle and the predicted result is found, the linkage prediction abnormality is established.
And if the loading specification abnormality and the linkage prediction abnormality are obtained, triggering a second early warning, wherein the second early warning is a higher-level warning for the loading behavior of the vehicle, reminding a supervisor to carry out detailed inspection and processing, and helping to strengthen the supervision on the loading of the vehicle and ensure the transportation safety and compliance.
The vehicle queue detection method comprises the steps of detecting a vehicle queue in the pre-test area, establishing vehicle queue detection feedback, generating a migration instruction if the vehicle queue detection feedback is that a passing vehicle queue result exists, and intelligently detecting the load of vehicles in the vehicle queue according to the migration instruction.
Specifically, in a pre-test area set in advance, a vehicle team is detected by a sensor, a camera, a radar, or the like. The motorcade detection is to monitor and analyze the real-time running state, the speed, the load and the like of a group of vehicles in a certain area, obtain the traffic state of the motorcade, establish the motorcade detection feedback and provide data support for subsequent processing.
And when the motorcade detection feedback indicates that the motorcade has a passing problem when passing through a certain area, namely, a passing motorcade result exists, generating a migration instruction. The traffic fleet result refers to whether the traffic conditions of the fleet on a certain road section are normal, and whether congestion or overlarge traffic flow occurs. The migration instruction is used for guiding the motorcade to reroute or adjusting the driving plan, so that the motorcade is prevented from being stopped or serious traffic accidents are avoided.
After the migration instruction is received, the vehicles in the motorcade are subjected to load detection, the load condition of each vehicle is measured, and whether the vehicles with overload or other abnormal loads exist or not can be intelligently judged by combining the factors such as the vehicle speed, road conditions and the like, so that the condition of non-compliant loads of the vehicles in the migration process is avoided.
The method further comprises the steps of performing early warning grade matching based on the first early warning and the second early warning, establishing an abnormal grade matching result, establishing an early warning response scheme by using the early warning grade matching result, and performing intelligent detection abnormal report according to the abnormal grade matching result and the early warning response scheme.
Specifically, the early warning grade matching is performed based on the first early warning and the second early warning, early warning information from different early warning sources is analyzed, grade classification is performed on the early warning information according to a certain rule, and an abnormal grade matching result is established. The first warning is typically a preliminary warning triggered based on abnormal loading behavior or load deflection values of the vehicle, while the second warning is a further warning based on abnormal loading specifications or linkage prediction anomalies. The aim of the early warning grade matching is to uniformly grade different types of early warning information so as to take corresponding measures according to different severity degrees. For example, if a first warning indicates that there is a slight overload of the vehicle and a second warning indicates that there is a serious risk of overload, the urgency of the problem is determined based on the result of the match of the two.
And establishing an early warning response scheme by using an early warning grade matching result, and making corresponding countermeasures. The early warning response scheme is determined based on the level of anomalies, high level anomalies may require immediate intervention, and low level anomalies may only require observation or subsequent inspection. For example, when a high level warning indicates that a vehicle is overloaded, the traffic police is instructed to check on site, or the vehicle is required to be temporarily stopped, while a low level warning may only require further detection of the vehicle at the next check.
And carrying out intelligent detection and abnormal report according to the abnormal grade matching result and the early warning response scheme, and reporting and responding to the abnormal condition according to the set early warning response scheme. If the detected abnormality meets a certain level, an alarm signal is sent out through an automatic means, and relevant personnel or departments are informed to process. For example, if a severe overload behavior of the vehicle is detected, an alarm will be immediately sent to inform traffic authorities to take necessary law enforcement measures.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the spirit or scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the present application and the equivalent techniques thereof, the present application is also intended to include such modifications and variations.