A method and system for generating a test report for a vehicle-road collaborative application scenario

Abstract

The present invention discloses a method and system for generating a test report for a vehicle-road cooperative application scenario, which relates to the field of intelligent transportation technology, including using sensors to collect vehicle speed, temperature, wind speed, and traffic flow data and performing preprocessing; extracting key features from the preprocessed data set to obtain vehicle condition characteristics; using a deep Q network to construct a behavior prediction model, inputting the vehicle condition characteristics into the behavior prediction model, and obtaining the vehicle's behavior prediction value in different scenarios. The present invention can form a closed-loop testing process from data collection to model prediction to actual testing. This closed-loop mechanism can not only improve the quality of data and the accuracy of the model, but also simulate the real traffic environment through mixed reality technology, thereby more accurately evaluating the performance of the vehicle-road cooperative system; in addition, by comparing and analyzing the prediction results with the actual test data, the performance of the behavior prediction model can be effectively evaluated, and the model and system design can be further optimized.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation technology, and in particular to a method and system for generating a vehicle-road collaborative application scenario test report. Background Art

[0002] In recent years, with the rapid development of intelligent transportation systems and autonomous driving technologies, vehicle-road cooperative technology has become one of the important research directions in the field of intelligent transportation. Vehicle-road cooperative technology aims to improve traffic safety, reduce traffic accidents, alleviate traffic congestion, and enhance the driving experience through information interaction between vehicles and road infrastructure. In this context, the method of generating test reports for vehicle-road cooperative application scenarios has become a key link in verifying and optimizing the performance of vehicle-road cooperative systems.

[0003] Existing vehicle-road collaboration test and evaluation methods mostly focus on the impact analysis of a single factor, only considering the vehicle's own sensor data or only evaluating the performance of a specific function, but fail to fully consider multiple environmental factors, including the impact of weather and traffic flow on vehicle behavior, resulting in a lack of comprehensiveness and accuracy in test results. In addition, most existing test methods rely on expert experience and simple statistical analysis to construct an evaluation system, lacking a systematic evaluation framework that can automatically adapt to different scenarios. Especially in the face of increasingly complex traffic environments and diversified needs, traditional methods are difficult to meet the growing test accuracy requirements. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a method and system for generating a test report for a vehicle-road cooperative application scenario, which solves the problem that existing vehicle-road cooperative test evaluation methods mostly focus on the impact analysis of a single factor, only consider the vehicle's own sensor data or only evaluate the performance of a specific function, but fail to fully consider a variety of environmental factors, including the impact of weather and traffic flow on vehicle behavior, resulting in a lack of comprehensiveness and accuracy in test results. In addition, most existing test methods rely on expert experience and simple statistical analysis to construct an evaluation system, and lack a systematic evaluation framework that can automatically adapt to different scenarios. Especially in the face of increasingly complex traffic environments and diversified needs, traditional methods are difficult to meet the problem of increasing test accuracy requirements.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a method for generating a vehicle-road cooperative application scenario test report, which includes:

[0008] S1: Use sensors to collect vehicle speed, temperature, wind speed, and traffic flow data and perform pre-processing;

[0009] S2: Extract key features from the preprocessed data set to obtain vehicle condition features;

[0010] S3: Use the deep Q network to build a behavior prediction model, input vehicle condition characteristics into the behavior prediction model, and obtain the vehicle's behavior prediction value in different scenarios;

[0011] S4: collect real-time data, input the real-time data into the behavior prediction model to obtain the real-time behavior prediction value, and compare and analyze the real-time behavior prediction value with the behavior prediction value;

[0012] S5: Calculate comprehensive evaluation indicators based on the comparative analysis results, evaluate the performance of the behavior prediction model, and generate a test report.

[0013] As a preferred solution of the method for generating a test report for a vehicle-road cooperative application scenario according to the present invention, wherein: the method of using sensors to collect vehicle speed, temperature, wind speed, and traffic flow data and performing preprocessing work, the specific steps are as follows:

[0014] Install speed sensors, temperature sensors, wind speed sensors, and traffic flow sensors on vehicles to collect data on vehicle speed, temperature, wind speed, and traffic flow;

[0015] Preprocess the collected data;

[0016] The pre-processed vehicle speed, temperature, wind speed and traffic flow data are integrated into the dataset, and the expression is:

[0017] x(t)=(v(t),T(t),W(t),F(t));

[0018] Among them, x(t) is the data set at time t, t represents the time node, v(t), T(t), W(t), and F(t) are the vehicle speed, ambient temperature, wind speed, and traffic flow at time t, respectively.

[0019] As a preferred solution of the method for generating a test report for a vehicle-road cooperative application scenario according to the present invention, the key features are extracted from the preprocessed data set to obtain vehicle condition features, specifically as follows:

[0020] Generate speed features based on vehicle speed data, the expression is:

[0021]

[0022] Where c1(t) represents the speed characteristic at time t; v(t+iΔt) represents the vehicle speed at time t+iΔt; is the average velocity in the time window, n is the number of samples in the time window, and Δt is the sampling interval;

[0023] Generate temperature features based on temperature data. The expression is:

[0024]

[0025] Where c2(t) is the temperature characteristic of the vehicle; T(t) represents the ambient temperature at time t; T ref is the reference temperature, T max and T min are the maximum and minimum temperature thresholds, respectively;

[0026] Generate wind speed characteristics based on wind speed data, the expression is:

[0027]

[0028] Where c3(t) is the wind speed characteristic of the vehicle; c3(t) represents the wind speed at time t; W max is the maximum wind speed threshold, α is the attenuation coefficient;

[0029] Generate traffic flow characteristics based on traffic flow data, the expression is:

[0030]

[0031] Among them, c4(t) represents the traffic flow characteristics of vehicles; F(t) represents the traffic flow at time t; F max is the maximum traffic flow threshold; β and γ are adjustment coefficients;

[0032] The speed characteristics, temperature characteristics, wind speed characteristics and traffic flow characteristics are integrated into the vehicle condition characteristics, and the expression is:

[0033] C(t)=[c1(t),c2(t),c3(t),c4(t)];

[0034] Among them, t is the time node, and C(t) represents the vehicle condition feature vector at time point t.

[0035] As a preferred solution of the method for generating a test report for a vehicle-road cooperative application scenario according to the present invention, the behavior prediction model is constructed using a deep Q network, specifically as follows:

[0036] Using deep Q-network as the basis of behavior prediction model;

[0037] Assume the state space is S, corresponding to the vehicle condition feature vector, the expression is:

[0038] S = {s1,s2,s3,s4};

[0039] Among them, s1 corresponds to c1(t); s2 corresponds to c2(t); s3 corresponds to c3(t); s4 corresponds to c4(t);

[0040] Assume that the action space is A, which contains three actions: acceleration, deceleration, and steering of the vehicle. The expression is:

[0041] A={a1,a2,a3};

[0042] Among them, a1 represents the acceleration action of the vehicle; a2 represents the deceleration action of the vehicle; a3 represents the steering action of the vehicle;

[0043] Construct a reward function R(s,a,s) based on the state space and action space ′ ), the expression is:

[0044] R(s,a,s ′ )=β1*exp(-α1*d1)+β2*exp(-α2*d2)+β3*exp(-α3*

[0045] d3);

[0046] Among them, s ′ represents the new state of the vehicle; d1 represents the distance to the nearest obstacle, d2 represents the distance to the preceding vehicle, and d3 represents the distance the vehicle deviates from the center of the lane; α1, α2, and α3 are the attenuation coefficients of d1, d2, and d3, respectively; β1, β2, and β3 correspond to the weight coefficients of d1, d2, and d3, respectively.

[0047] As a preferred solution of the method for generating a test report for a vehicle-road cooperative application scenario according to the present invention, the vehicle condition characteristics are input into the behavior prediction model to obtain the predicted behavior value of the vehicle in different scenarios, which is expressed as follows:

[0048] P(t)=argmax a Q(s,a;θ);

[0049] Among them, P(t) is the behavior prediction value, θ is the parameter vector, argmax a It means finding the best action among all possible actions a.

[0050] As a preferred solution of the method for generating a test report for a vehicle-road cooperative application scenario according to the present invention, the method collects real-time data, inputs the real-time data into a behavior prediction model, obtains a real-time behavior prediction value, and compares and analyzes the real-time behavior prediction value with the behavior prediction value, as follows:

[0051] Collect real-time data, process the real-time data to obtain real-time vehicle condition characteristics, and input the real-time vehicle condition characteristics into the behavior prediction model to obtain the real-time behavior prediction value Ptest(t);

[0052] Create a comparative analysis function H based on the recorded data and the prediction results of the behavior prediction model. The expression is:

[0053] H = n1∑i = 1n(Ptest(t + iΔt)∣Ptest(t + iΔt) - P(t + iΔt)∣)2·

[0054] exp(-αo∣Δv(t + iΔt)∣);

[0055] Where, Ptest(t) represents the real-time behavior prediction value, n represents the number of samples, Δt is the sampling interval time, αo is the attenuation coefficient, Δv represents the speed change amount, Δv(t + iΔt) is the speed change amount at time t + iΔt, i represents the index variable, and exp represents the exponential function.

[0056] As a preferred solution of the method for generating a test report on the vehicle-road collaborative application scenario described in this invention, wherein: calculating a comprehensive evaluation index based on the comparative analysis result, evaluating the performance of the behavior prediction model, and generating a test report, specifically as follows:

[0057] Create a comprehensive evaluation index E(t) based on the result of the comparative analysis. The expression is:

[0058]

[0059] Where, j represents the index of each time point in the time series; j iterates from 1 to m, m is the total number of sampling points, and Δt is the sampling interval time;

[0060] The value of E(t) ranges from 0 to positive infinity;

[0061] Set a threshold TT;

[0062] When 0 < E(t) ≤ TT, it indicates that the prediction ability of the behavior prediction model is strong;

[0063] When E(t) > threshold TT, it indicates that the prediction ability of the behavior prediction model is poor, and the data is re-input into the behavior prediction model for training until 0 < E(t) ≤ TT;

[0064] Generate a test report based on the result of E(t). The test report includes the specific value of E(t), the basic information of the model, the time period and conditions of data collection, the type of sensors used and their accuracy, the description of the experimental environment, the detailed results of the comparative analysis, suggestions for model optimization, and future improvement directions.

[0065] In a second aspect, the present invention provides a vehicle-road cooperative application scenario test report generation system, comprising:

[0066] The data acquisition module is responsible for collecting vehicle speed, temperature, wind speed, and traffic flow data using sensors. This module also includes preprocessing processes such as data cleaning, missing value processing, and outlier detection;

[0067] The feature extraction module is responsible for extracting key features from the preprocessed data set and extracting vehicle condition features from the key features;

[0068] The prediction module is responsible for building a behavior prediction model using a deep Q network and inputting vehicle condition characteristics into the behavior prediction model to obtain the vehicle's behavior prediction values ​​in different scenarios;

[0069] The analysis module is responsible for inputting real-time data into the behavior prediction model to obtain real-time behavior prediction values, and performing comparative analysis between the real-time behavior prediction values ​​and the behavior prediction values;

[0070] The performance evaluation module is responsible for calculating comprehensive evaluation indicators using comparative analysis results, evaluating the performance of the behavior prediction model, and organizing the test data to generate an easy-to-understand test report.

[0071] In a third aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the method for generating a vehicle-road collaborative application scenario test report as described in the first aspect of the present invention.

[0072] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium on which a computer program is stored, wherein: when the computer program is executed by a processor, it implements any step of the method for generating a vehicle-road collaborative application scenario test report as described in the first aspect of the present invention.

[0073] The beneficial effects of the present invention are: the present invention can form a closed-loop testing process from data collection to model prediction to actual testing. This closed-loop mechanism can not only improve the quality of data and the accuracy of the model, but also simulate the real traffic environment through mixed reality technology, thereby more accurately evaluating the performance of the vehicle-road cooperative system; in addition, by comparing and analyzing the prediction results with the actual test data, the performance of the behavior prediction model can be effectively evaluated, the model and system design can be further optimized, the efficiency and accuracy of the vehicle-road cooperative scenario test can be significantly improved, and the development and improvement of the intelligent transportation system can be promoted. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. 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 any creative work.

[0075] Figure 1 This is a flowchart of the method for generating a test report for a vehicle-road collaborative application scenario in Example 1.

[0076] Figure 2 This is a flowchart of the vehicle-road collaborative application scenario test report generation system in Example 1. DETAILED DESCRIPTION

[0077] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0078] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0079] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0080] Example 1, with reference to Figure 1 and Figure 2 , which is the first embodiment of the present invention, provides a method for generating a vehicle-road cooperative application scenario test report, comprising the following steps:

[0081] S1 uses sensors to collect vehicle speed, temperature, wind speed, and traffic flow data and performs pre-processing work;

[0082] Install speed sensors, temperature sensors, wind speed sensors, and traffic flow sensors on vehicles to collect data on vehicle speed, temperature, wind speed, and traffic flow;

[0083] The precise capture of multi-dimensional data during vehicle driving provides a high-quality data foundation for subsequent data analysis.

[0084] Use data preprocessing algorithms to preprocess the collected data. The preprocessing work includes:

[0085] Data cleaning: remove invalid data points and noise data to ensure data validity;

[0086] Missing value processing: interpolation method is used to fill missing values. Here, linear interpolation method is used. That is, for missing data points, linear interpolation is performed using the nearest valid data points before and after to estimate the missing values.

[0087] Outlier detection: Statistical methods are used to identify and process outliers. Specifically, the Z-score method is used, that is, any value below the mean minus 3 standard deviations or above the mean plus 3 standard deviations is considered an outlier, and the outlier is replaced by the mean of the variable.

[0088] The pre-processed vehicle speed, temperature, wind speed and traffic flow data are integrated into the dataset, and the expression is:

[0089] x(t)=(v(t),T(t),W(t),F(t));

[0090] Among them, x(t) is the data set at time t, t represents the time node, v(t), T(t), W(t), and F(t) are the vehicle speed, ambient temperature, wind speed, and traffic flow at time t, respectively.

[0091] Preprocessing ensures the consistency and integrity of data, thereby improving the reliability and effectiveness of the entire system.

[0092] S2 extracts key features from the preprocessed data set to obtain vehicle condition features;

[0093] Based on the vehicle speed data, the speed feature is generated using the standard deviation calculation method in statistics. The expression is:

[0094]

[0095] Where c1(t) represents the speed characteristic at time t; v(t+iΔt) represents the vehicle speed at time t+iΔt; is the average velocity in the time window, n is the number of samples in the time window, and Δt is the sampling interval;

[0096] Based on the temperature data, the normalization method is used to generate the temperature feature, which is expressed as:

[0097]

[0098] Where c2(t) is the temperature characteristic of the vehicle; T(t) represents the ambient temperature at time t; T ref is the reference temperature, T max and T minare the maximum and minimum temperature thresholds, respectively;

[0099] Based on the wind speed data, the wind speed characteristics are generated by combining linear normalization and exponential decay methods. The expression is:

[0100]

[0101] Where c3(t) is the wind speed characteristic of the vehicle; c3(t) represents the wind speed at time t; W max is the maximum wind speed threshold, α is the attenuation coefficient;

[0102] Based on traffic flow data, the traffic flow characteristics are generated by combining linear normalization and periodic function modulation. The expression is:

[0103]

[0104] Among them, c4(t) represents the traffic flow characteristics of vehicles; F(t) represents the traffic flow at time t; F max is the maximum traffic flow threshold; β and γ are adjustment coefficients;

[0105] The speed characteristics, temperature characteristics, wind speed characteristics and traffic flow characteristics are integrated into the vehicle condition characteristics, and the expression is:

[0106] C(t)=[c1(t),c2(t),c3(t),c4(t)];

[0107] Where t is the time node, C(t) represents the vehicle condition feature vector at time point t;

[0108] Specifically, speed characteristics refer to the degree of speed fluctuation within time t, temperature characteristics refer to the factor that affects the degree of ambient temperature on vehicle performance, wind speed characteristics refer to the factor that measures the degree of influence of wind speed on vehicle driving, and traffic flow characteristics refer to the degree of influence of traffic flow on vehicle driving.

[0109] By extracting key features, not only the data structure is simplified, but also the correlation between features is enhanced, enabling the behavior prediction model to more effectively learn and understand the vehicle's performance under different environmental conditions, further improving the model's ability to understand complex scenarios and prediction accuracy.

[0110] S3 uses a deep Q network to build a behavior prediction model, inputs vehicle condition characteristics into the behavior prediction model, and obtains the vehicle's behavior prediction value in different scenarios;

[0111] Using deep Q-network as the basis of behavior prediction model;

[0112] Assume the state space is S, corresponding to the vehicle condition feature vector, the expression is:

[0113] S = {s1,s2,s3,s4};

[0114] Among them, s1 corresponds to c1(t); s2 corresponds to c2(t); s3 corresponds to c3(t); s4 corresponds to c4(t);

[0115] Assume that the action space is A, which contains three actions: acceleration, deceleration, and steering of the vehicle. The expression is:

[0116] A={a1,a2,a3};

[0117] Among them, a1 represents the acceleration action of the vehicle; a2 represents the deceleration action of the vehicle; a3 represents the steering action of the vehicle;

[0118] Construct a reward function R(s,a,s) based on the state space and action space ′ ), the expression is:

[0119] R(s,a,s ′ )=β1*exp(-α1*d1)+β2*exp(-α2*d2)+β3*exp(-α3*

[0120] d3);

[0121] Among them, s ′ represents the new state of the vehicle after executing a; d1 represents the distance to the nearest obstacle, d2 represents the distance to the preceding vehicle, and d3 represents the distance the vehicle deviates from the center of the lane; α1, α2, and α3 are the attenuation coefficients of d1, d2, and d3, respectively; β1, β2, and β3 correspond to the weight coefficients of d1, d2, and d3, respectively.

[0122] The reward function works as follows:

[0123] Guided behavior: The reward function guides the vehicle’s behavior by giving positive or negative reward values. Positive rewards encourage the vehicle to take similar behaviors, while negative rewards prevent the vehicle from repeating the same behavior.

[0124] Optimizing strategy: By maximizing the cumulative reward, the reinforcement learning algorithm can learn the best behavior strategy. The design of the reward function directly affects what strategy the agent learns.

[0125] Safety considerations: The reward function can be designed to include safety considerations, such as reducing the distance to the obstacle will receive a negative reward, while maintaining an appropriate safety distance will receive a positive reward.

[0126] Efficiency and comfort: In addition to safety, the reward function can also consider driving efficiency (such as reducing unnecessary acceleration and deceleration) and passenger comfort (such as smooth acceleration and deceleration).

[0127] Multi-objective optimization: In complex driving environments, the reward function may need to consider multiple objectives simultaneously, such as time to destination, fuel consumption, passenger satisfaction, etc. The design of the reward function needs to balance these different objectives.

[0128] S4 collects real-time data, inputs the real-time data into the behavior prediction model, obtains the real-time behavior prediction value, and compares and analyzes the real-time behavior prediction value with the behavior prediction value;

[0129] The expression of the prediction model is:

[0130] P(t)=argmax a Q(s,a;θ);

[0131] Among them, P(t) is the behavior prediction value, θ is the parameter vector, argmax a It means finding the best action among all possible actions a;

[0132] Collect and process real-time data to obtain real-time vehicle condition characteristics, input the real-time vehicle condition characteristics into the behavior prediction model, and obtain the real-time behavior prediction value Ptest(t);

[0133] A comparative analysis function H is created based on the recorded data and the prediction results of the behavior prediction model. The expression is:

[0134] H=n1∑i=1n(Ptest(t+iΔt)∣Ptest(t+iΔt)-P(t+iΔt)∣)2·

[0135] exp(-αo|Δv(t+iΔt)|);

[0136] Where Ptest(t) represents the real-time behavior prediction value, n represents the number of samples, Δt is the sampling interval, αo is the attenuation coefficient, Δv represents the speed change, Δv(t+iΔt) is the speed change at time t+iΔt, i represents the index variable, and exp represents the exponential function.

[0137] S5 calculates comprehensive evaluation indicators based on comparative analysis results, evaluates the performance of the behavior prediction model, and generates a test report;

[0138] Based on the results of comparative analysis, a comprehensive evaluation index E(t) is created, which is expressed as:

[0139]

[0140] Where j represents the index of each time point in the time series; j iterates from 1 to m, m is the total number of sampling points, and Δt is the sampling interval;

[0141] The value range of E(t) is from 0 to positive infinity;

[0142] Set a threshold TT, and the value of TT can be flexibly adjusted according to specific test requirements;

[0143] When 0 < E(t) ≤ TT, it indicates that the prediction ability of the behavior prediction model is strong;

[0144] When E(t) > the threshold TT, it indicates that the prediction ability of the behavior prediction model is poor. The data will be re - input into the behavior prediction model for training until 0 < E(t) ≤ TT.

[0145] The smaller the evaluation index E(t), the higher the prediction accuracy of the model. Through this method, we can quantitatively understand the performance of the model at different times and in different scenarios, and further optimize the model accordingly;

[0146] Based on the results of E(t), generate a test report. The test report includes the specific value of E(t), the basic information of the model, the time period and conditions of data collection, the type of sensors used and their accuracy, the description of the experimental environment, the detailed results of comparative analysis, suggestions for model optimization, and future improvement directions.

[0147] This embodiment also provides a test report generation system for vehicle - road collaborative application scenarios, including:

[0148] A data acquisition module, responsible for collecting vehicle driving speed, temperature, wind speed, and traffic flow data using sensors. This module also includes a pre - processing process of data cleaning, missing value handling, and outlier detection;

[0149] A feature extraction module, responsible for extracting key features from the pre - processed dataset and extracting vehicle condition features from the key features;

[0150] A prediction module, responsible for constructing a behavior prediction model using a deep Q - network and inputting the vehicle condition features into the behavior prediction model to obtain the behavior prediction values of the vehicle in different scenarios;

[0151] An analysis module, responsible for inputting real - time data into the behavior prediction model to obtain real - time behavior prediction values, and comparing and analyzing the real - time behavior prediction values with the behavior prediction values;

[0152] Responsible for calculating a comprehensive evaluation index using the comparative analysis results, evaluating the performance of the behavior prediction model, and organizing the test data to generate an easy - to - understand test report.

[0153] This embodiment also provides a computer device, which is suitable for the method of generating a test report for a vehicle-road collaborative application scenario, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the method of generating a test report for a vehicle-road collaborative application scenario proposed in the above embodiment.

[0154] The computer device may be a terminal, comprising a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.

[0155] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by the processor, it implements the method for generating a test report for a vehicle-road cooperative application scenario proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk.

[0156] In summary, the present invention forms a closed-loop testing process from data collection to model prediction to actual testing. This closed-loop mechanism can not only improve the quality of data and the accuracy of the model, but also simulate the real traffic environment through mixed reality technology, thereby more accurately evaluating the performance of the vehicle-road cooperative system; in addition, by comparing and analyzing the prediction results with the actual test data, the performance of the behavior prediction model can be effectively evaluated, and the model and system design can be further optimized, which can significantly improve the efficiency and accuracy of the vehicle-road cooperative scenario test and promote the development and improvement of the intelligent transportation system.

[0157] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for generating a test report for a vehicle-road collaboration application scenario, characterized by: include, S1: Use sensors to collect vehicle speed, temperature, wind speed, and traffic flow data and perform pre-processing; S2: Extract key features from the preprocessed dataset to obtain vehicle condition features, as follows: Generate speed features based on vehicle speed data, the expression is: ； in, Indicates time Velocity characteristics within; Indicates time Vehicle speed; is the average velocity in the time window, is the number of samples in the time window, is the sampling interval; Generate temperature features based on temperature data. The expression is: ； in, is the temperature characteristic of the vehicle; Indicates time ambient temperature; is the reference temperature, and are the maximum and minimum temperature thresholds, respectively; Generate wind speed characteristics based on wind speed data, the expression is: ； in, is the wind speed characteristic of the vehicle; Indicates time wind speed; is the maximum wind speed threshold, is the attenuation coefficient, Indicates time wind speed; Generate traffic flow characteristics based on traffic flow data, the expression is: ； in, Indicates the traffic flow characteristics of vehicles; Indicates time Traffic flow; is the maximum traffic flow threshold; and is the adjustment coefficient; The speed characteristics, temperature characteristics, wind speed characteristics and traffic flow characteristics are integrated into the vehicle condition characteristics, and the expression is: ； in, It's a time node. Indicates at a point in time Vehicle condition feature vector; S3: Use the deep Q network to build a behavior prediction model. Input the vehicle condition characteristics into the behavior prediction model to obtain the vehicle's behavior prediction values ​​in different scenarios. The details are as follows: Using deep Q-network as the basis of behavior prediction model; Set the state space to , corresponding to the vehicle condition feature vectors respectively; Set the action space to , which includes three actions: acceleration, deceleration and steering of the vehicle; Constructing reward function based on state space and action space , the expression is: ； in, Indicates the new state of the vehicle; Indicates the distance to the nearest obstacle. Indicates the distance to the vehicle ahead. Indicates the distance the vehicle deviates from the center of the lane; 、 and They are 、 and The attenuation coefficient; 、 and Corresponding respectively 、 and The weight coefficient of S4: Collect real-time data, input the real-time data into the behavior prediction model, obtain the real-time behavior prediction value, and compare and analyze the real-time behavior prediction value with the behavior prediction value, as follows: Collect real-time data, process the real-time data, obtain real-time vehicle condition characteristics, input the real-time vehicle condition characteristics into the behavior prediction model, and obtain the real-time behavior prediction value ; Create a comparative analysis function based on the recorded data and the prediction results of the behavior prediction model , the expression is: ； in, represents the real-time behavior prediction value, represents the number of samples, is the sampling interval, is the attenuation coefficient, Indicates the speed change, It's in time The velocity change at Represents an index variable, represents the exponential function; S5: Calculate comprehensive evaluation indicators based on the comparative analysis results, evaluate the performance of the behavior prediction model, and generate a test report.

2. The method for generating a vehicle-road cooperative application scenario test report according to claim 1, characterized in that: The method uses sensors to collect vehicle speed, temperature, wind speed, and traffic flow data and performs preprocessing work, and the specific steps are as follows: Install speed sensors, temperature sensors, wind speed sensors, and traffic flow sensors on vehicles to collect data on vehicle speed, temperature, wind speed, and traffic flow; Preprocess the collected data; The pre-processed vehicle speed, temperature, wind speed and traffic flow data are integrated into the dataset.

3. The method for generating a vehicle-road cooperative application scenario test report according to claim 2, characterized in that: The vehicle condition characteristics are input into the behavior prediction model to obtain the behavior prediction values ​​of the vehicle in different scenarios.

4. The method for generating a vehicle-road cooperative application scenario test report according to claim 3, characterized in that: The comprehensive evaluation index is calculated based on the comparative analysis results to evaluate the performance of the behavior prediction model and generate a test report, as follows: Create comprehensive evaluation indicators based on the results of comparative analysis ; The value of is from 0 to positive infinity; Set the threshold TT; when When ≤TT, it indicates that the behavior prediction model has strong predictive ability; when When the threshold TT is >, it indicates that the prediction ability of the behavior prediction model is poor. The data is re-input into the behavior prediction model for training until ≤TT; based on The test report includes The specific values ​​of , basic information about the model, the time period and conditions for data collection, the type of sensors used and their accuracy, a description of the experimental environment, detailed results of the comparative analysis, suggestions for model optimization, and future improvement directions.

5. A vehicle-road cooperative application scenario test report generation system, used to implement the vehicle-road cooperative application scenario test report generation method according to any one of claims 1 to 4, characterized in that: include: The data acquisition module is responsible for collecting vehicle speed, temperature, wind speed, and traffic flow data using sensors. This module also includes preprocessing processes such as data cleaning, missing value processing, and outlier detection; The feature extraction module is responsible for extracting key features from the preprocessed data set and extracting vehicle condition features from the key features; The prediction module is responsible for building a behavior prediction model using a deep Q network and inputting vehicle condition characteristics into the behavior prediction model to obtain the vehicle's behavior prediction values ​​in different scenarios; The analysis module is responsible for inputting real-time data into the behavior prediction model to obtain real-time behavior prediction values, and performing comparative analysis between the real-time behavior prediction values ​​and the behavior prediction values; The performance evaluation module is responsible for calculating comprehensive evaluation indicators using comparative analysis results, evaluating the performance of the behavior prediction model, and organizing the test data to generate an easy-to-understand test report.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for generating a vehicle-road cooperative application scenario test report as described in any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for generating a vehicle-road cooperative application scenario test report as described in any one of claims 1 to 4 are implemented.

Images