Unmanned vehicle comprehensive quantitative evaluation test road system and method thereof

Through the comprehensive quantitative evaluation and testing of road systems of unmanned vehicles, multiple sensors and AI/ML algorithms are used to evaluate the adaptability and execution accuracy of vehicles in different road environments, and solve the problem that the existing technology cannot fully reflect the comprehensive performance of vehicles. , and achieved detailed classification evaluation and performance improvement.

CN119086085APending Publication Date: 2024-12-06SHENYANG YUANSONG TECHNOLOGY CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410891223.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-04
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The existing comprehensive quantitative evaluation technology for unmanned vehicles cannot comprehensively, systematically and objectively reflect the vehicle's adaptability and execution accuracy under different road conditions, and cannot conduct detailed and scientific classification evaluation.

Method used

A comprehensive quantitative evaluation and testing road system for driverless vehicles is proposed, including vehicle information collection module, road adaptation division module, evaluation result evaluation module and summary evaluation module. Data is collected in real time through multiple sensors and AI/ML algorithms to divide different road environment scenarios , evaluate the safety, intelligence, stability and interaction of vehicles, and conduct comprehensive evaluation.

Benefits of technology

A detailed and scientific classification evaluation of driverless vehicles in different road environments is achieved, analyzing the advantages and disadvantages of vehicles is provided, and the comprehensiveness and objectivity of test evaluation is improved, which helps improve vehicle performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119086085A_ABST
    Figure CN119086085A_ABST
Patent Text Reader

Abstract

The invention provides an unmanned vehicle comprehensive quantitative evaluation test road system and method, and the system comprises a vehicle information collection module, a road adaptation division module, an evaluation result evaluation module, and a summary evaluation module. According to the method, scenes with different difficulties formed by different environmental elements in a suburban driving environment and an urban driving environment are divided, the execution accuracy of road conditions is analyzed in a graded manner, data are provided for intelligent identification and execution behavior evaluation in a driving task process by taking a judgment principle as a reference, accurate evaluation is carried out, and the driving task efficiency is improved. Different road scenes in the road adaptation division module are used for difficulty division and grading, judgment is carried out by taking the evaluation result evaluation module as a principle, and the execution accuracy and accuracy of the unmanned vehicle in different difficulty environments are distinguished and evaluated, so that detailed and scientific classification evaluation is realized. The comprehensive performance of unmanned driving is reflected comprehensively, systematically and objectively.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to unmanned driving related fields, and in particular to an unmanned driving vehicle comprehensive quantitative evaluation test road system and method. Background Art

[0002] Road testing needs to consider a variety of factors, such as road width, road flatness, and traffic congestion, and conduct a comprehensive analysis to make a comprehensive judgment and evaluation of the driving mission of the unmanned vehicle. However, under the existing comprehensive quantitative evaluation of unmanned vehicles, the adaptability and execution accuracy of unmanned vehicles under different road conditions have not been carefully and scientifically classified and evaluated, and cannot comprehensively, systematically and objectively reflect the comprehensive performance of unmanned vehicles.

[0003] Therefore, in order to comprehensively, systematically and objectively reflect the comprehensive performance of unmanned vehicles, it is necessary to improve and refine the analysis of the adaptability and execution accuracy of unmanned vehicles under different road conditions, and analyze the advantages and disadvantages of unmanned vehicles, so as to standardize the research on comprehensive quantitative evaluation technology of intelligent driving vehicles and promote the improvement of relevant test and evaluation systems. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a comprehensive quantitative evaluation test road system and method for unmanned vehicles.

[0005] Based on this, the present invention proposes a comprehensive quantitative evaluation test road system and method for unmanned vehicles, including a vehicle information collection module, a road adaptation division module, an evaluation result evaluation module, and a summary evaluation module;

[0006] The vehicle information collection module is used to collect, record and transmit various types of data in real time or offline, and control path planning and execution behavior;

[0007] The road adaptation division module is divided into scenarios of different difficulty composed of different environmental factors in suburban driving environment and urban driving environment, providing data for the vehicle's optimal adaptation range and accuracy evaluation under the scenario;

[0008] The evaluation result evaluation module provides data for intelligent recognition and execution behavior evaluation during vehicle driving tasks based on the judgment principle;

[0009] The summary evaluation module, the comprehensive road adaptation division module, and the test result evaluation module provide a comprehensive evaluation of vehicle performance and provide improvement data references for vehicle disadvantages.

[0010] Optionally, the road adaptation division module mainly comprises dynamic factors, traffic signs, real-time weather and road conditions for different environmental elements in suburban driving environment and urban driving environment, and obtains the complexity of real-time road conditions by analyzing and modeling the road condition data.

[0011] Optionally, include the following steps:

[0012] S1, the vehicle information collection module uses a variety of sensors and computer equipment on the body of the unmanned vehicle to collect, record and transmit various types of data in real time or offline, as well as control path planning and execution behavior;

[0013] S2, using the road adaptation classification module to classify the test scenarios of the unmanned vehicle. The module is divided into scenarios composed of different environmental elements under the suburban driving environment unit and the urban driving environment unit, and the road conditions for vehicle testing are divided and graded;

[0014] S3, using the evaluation result evaluation module, based on the four sets of judgment principles, evaluates the accuracy of the vehicle's intelligent recognition and execution behavior during the driving task under different road conditions;

[0015] S4, the summary evaluation module weights the overall detection of the unmanned vehicle according to the vehicle information collection module, the road adaptation division module, and the evaluation result evaluation module according to the sequence relationship analysis method, and uses the comprehensive evaluation method to comprehensively evaluate the unmanned vehicle to obtain the final evaluation.

[0016] Optionally, in step S1, a variety of sensors are used to identify traffic signs, vehicles, road conditions, and real-time weather, and computer equipment is used to organize data and build maps. Through AI and ML algorithms, driverless cars can identify road conditions, traffic rules, and other road users, and predict their behavior. Based on this information, the vehicle will make real-time decisions, such as driving speed, driving direction, lane changing, overtaking, avoiding pedestrians and other vehicles, and control the path planning and execution behavior of the driverless vehicle.

[0017] Optionally, in step S2, the suburban driving environment unit and the urban driving environment unit both include elements such as dynamic vehicles and pedestrians, traffic road signs, real-time weather, road conditions, etc.;

[0018] The dynamic vehicles and pedestrians are the number of vehicles and pedestrians, which are reflected in the degree of road congestion and the interaction of unmanned vehicles. The visual sensors on the unmanned vehicles collect video data from the vehicle monitoring equipment, and extract vehicle and pedestrian information based on the data to determine the degree of road congestion. The difficulty of the road conditions is identified and divided into difficult / medium / easy levels; traffic roads and signs are divided into many / medium / few levels according to the frequency of their appearance; real-time weather includes sunny days, rainy days, snowy days, foggy days and sand and dust elements. The real-time weather is explored through the fusion solution of laser radar, camera and millimeter wave radar. Through combined use, the situation of laser radar detection errors in snowy days is avoided. The excellent penetration ability of millimeter wave radar in rain, snow and fog is used to achieve accurate detection and avoid the failure of positioning of unmanned driving lane lines; road conditions include the width of the road, the unevenness of the road surface and the construction status of the road surface, which are divided into difficult / medium / easy levels. The road condition levels are classified and distinguished by environmental elements of different difficulty levels, which is convenient for judging the comprehensive performance of unmanned vehicles in different environmental roads.

[0019] Optionally, in step S3, the judgment principles include safety, intelligence, stability and interactivity;

[0020] Safety includes the functions, appearance and speed of the driverless vehicle; intelligence is reflected in the external environment, the judgment of traffic signal signs and driving tasks, the perception of the surrounding environment through Autoware's algorithm, the detection of moving vehicles and pedestrians, the avoidance of traffic accidents, and the detection of traffic lights and traffic police gesture recognition to avoid violating traffic rules; stability includes speed changes and steering return during driving; interactivity is used for communication between the vehicle and the user;

[0021] The safety scoring criteria include detecting whether the functions of the driverless vehicle are intact after a collision, whether the driverless vehicle is damaged after completing the driving mission, and whether the driverless vehicle drives at the prescribed speed under bad weather and road rules, and whether it provides users with a comfortable experience;

[0022] The intelligence of the driverless vehicle is systematically scored through signal recognition, traffic police gesture recognition, pedestrian and non-motor vehicle recognition accuracy, appropriateness of vehicle distance maintenance, rationality of route changes, and correct recognition and reasonable imitation of special lane giving. Special lane giving refers to making reasonable way for ambulances during emergency rescue missions. The sound of the ambulance is judged by the Autoware algorithm, and the behavior of the front and rear vehicles is detected by the video surveillance system, and reasonable imitation of lane giving is implemented. The stability is scored by whether the vehicle speed changes smoothly throughout the driving process and whether the steering response speed is reasonable. Finally, the interactivity is scored based on the interactive response between the driverless vehicle and the user.

[0023] Optionally, the intelligence includes signal signs, pedestrian and non-motor vehicle recognition, vehicle distance keeping, traffic police gesture recognition, route change, and special way-giving. Among them, route change obtains satellite images through aerospace technology, and then uses intelligent algorithms to extract information such as road congestion to avoid bad road conditions.

[0024] Optionally, in step S4, when the summary evaluation module conducts a comprehensive evaluation of the drone driving vehicle, the difficulty level is divided according to the different road scenarios in the road adaptation division module, and the judgment is made based on the principles in the evaluation result evaluation module. The execution accuracy and accuracy of the unmanned vehicle in different difficulty environments are evaluated, the adaptability range of the unmanned vehicle on different roads is determined, and comprehensive calculations are used to analyze the shortcomings and disadvantages of the unmanned vehicle to provide data reference for improvement.

[0025] The present invention has the following advantages: The present invention provides a comprehensive quantitative evaluation test road system and method for unmanned vehicles through improvement, which has the following improvements compared with similar devices:

[0026] Advantage 1: The present invention provides a comprehensive quantitative evaluation test road system and method for unmanned vehicles. By setting up a road adaptation division module, different difficulty scenarios composed of different environmental elements in suburban driving environments and urban driving environments are divided, providing data for the vehicle's optimal adaptation range and the accuracy evaluation of execution in the scenario, and realizing accurate analysis of the execution accuracy of road conditions in a graded manner.

[0027] Advantage 2: The present invention provides a comprehensive quantitative evaluation test road system and method for unmanned vehicles. By setting up an evaluation result evaluation module, it uses four major judgment principles as a benchmark, namely, safety, intelligence, stability and interactivity to provide data for the intelligent recognition and execution behavior evaluation of the vehicle during the driving task, and conducts accurate evaluation.

[0028] Advantage 3: The present invention provides a comprehensive quantitative evaluation test road system and method for unmanned vehicles. A summary evaluation module is set up. When the summary evaluation module conducts a comprehensive evaluation of the unmanned vehicle, the difficulty level is divided according to the different road scenarios in the road adaptation division module, and the judgment is made according to the evaluation result evaluation module. The two are combined with each other to differentiate and evaluate the execution accuracy and accuracy of the unmanned vehicle in different difficulty environments, summarize the advantages and disadvantages of the unmanned vehicle, provide data reference for subsequent improvements, realize detailed and scientific classification evaluation, and comprehensively, systematically and objectively reflect the comprehensive performance of the unmanned driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is the overall flow chart of the present invention;

[0030] Figure 2 This is a flow chart of the road adaptation division module of the present invention;

[0031] Figure 3 This is a flow chart of the evaluation module of the test result of the present invention;

[0032] Figure 4 The evaluation module flow chart is summarized for the present invention. DETAILED DESCRIPTION

[0033] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0034] Embodiment 1

[0035] Please refer to Figure 1 , this embodiment provides a comprehensive quantitative evaluation test road system for unmanned vehicles, including a vehicle information collection module, a road adaptation division module, an evaluation result evaluation module, and a summary evaluation module; the vehicle information collection module is used to collect, record and transmit various types of data in real time or offline, as well as control path planning and execution behavior; the road adaptation division module is divided into scenarios of different difficulty composed of different environmental elements in suburban driving environment and urban driving environment, providing data for the evaluation of the best adaptation range of the vehicle and the accuracy of execution in the scenario; the evaluation result evaluation module, based on the judgment principle, provides data for the intelligent recognition and execution behavior evaluation during the vehicle driving task; the summary evaluation module, the comprehensive road adaptation division module and the test result evaluation module provide a comprehensive evaluation of vehicle performance, and provide improvement data reference for vehicle disadvantages, the road adaptation division module is mainly composed of dynamic factors, traffic signs, real-time weather and road conditions for different environmental elements in suburban driving environment and urban driving environment, and the complexity of real-time road conditions is obtained by analyzing and modeling the road condition data;

[0036] The vehicle information collection module uses a variety of sensors and computer equipment installed on the body of the driverless vehicle to collect, record and transmit various types of data in real time or offline, and uses computer equipment to organize and map the data. Through AI and ML algorithms, driverless cars can identify road conditions, traffic rules and other road users, and predict their behavior. Based on this information, the vehicle will make real-time decisions, such as driving speed, driving direction, lane changing, overtaking, avoiding pedestrians and other vehicles, and control the path planning and execution behavior of the driverless vehicle.

[0037] Embodiment 2

[0038] Please refer to Figure 2-4This embodiment provides a comprehensive quantitative evaluation test method for an unmanned vehicle, comprising the following steps:

[0039] S1, the vehicle information collection module uses a variety of sensors and computer equipment on the body of the unmanned vehicle to collect, record and transmit various types of data in real time or offline, as well as control path planning and execution behavior;

[0040] S2, using the road adaptation classification module to classify the test scenarios of the unmanned vehicle. The module is divided into scenarios composed of different environmental elements under the suburban driving environment unit and the urban driving environment unit, and the road conditions for vehicle testing are divided and graded;

[0041] S3, using the evaluation result evaluation module, based on the four sets of judgment principles, evaluates the accuracy of the vehicle's intelligent recognition and execution behavior during the driving task under different road conditions;

[0042] S4, the summary evaluation module weights the overall detection of the unmanned vehicle according to the vehicle information collection module, the road adaptation division module, and the evaluation result evaluation module according to the sequence relationship analysis method, and uses the comprehensive evaluation method to comprehensively evaluate the unmanned vehicle to obtain the final evaluation;

[0043] In step S1, a variety of sensors are used to identify traffic signs, vehicles, road conditions, and real-time weather, and computer equipment is used to organize data and build maps. Through AI and ML algorithms, driverless cars can identify road conditions, traffic rules, and other road users, and predict their behavior. Based on this information, the vehicle will make real-time decisions, such as driving speed, driving direction, lane changing, overtaking, avoiding pedestrians and other vehicles, and control the path planning and execution of the driverless vehicle.

[0044] In step S2, the suburban driving environment unit and the urban driving environment unit both include dynamic vehicles and pedestrians, traffic road signs, real-time weather, road conditions and other elements;

[0045] The dynamic vehicles and pedestrians are the number of vehicles and pedestrians, which are reflected in the degree of road congestion and the interaction of unmanned vehicles. The visual sensors on the unmanned vehicles collect video data from the vehicle monitoring equipment, and extract vehicle and pedestrian information based on the data to determine the degree of road congestion. The difficulty of the road conditions is identified and divided into difficult / medium / easy levels; traffic roads and signs are divided into many / medium / few levels according to the frequency of their appearance; real-time weather includes sunny days, rainy days, snowy days, foggy days and sand and dust elements. The real-time weather is explored through the fusion solution of laser radar, camera and millimeter wave radar. Through combined use, the situation of laser radar detection errors in snowy days is avoided. The excellent penetration ability of millimeter wave radar in rain, snow and fog is used to achieve accurate detection and avoid the failure of positioning of unmanned driving lane lines; road conditions include the width of the road, the unevenness of the road surface and the construction status of the road surface, which are divided into difficult / medium / easy levels. The road condition levels are classified and distinguished by environmental elements of different difficulty levels, which is convenient for judging the comprehensive performance of unmanned vehicles in different environmental roads.

[0046] In step S3, the judgment principles include safety, intelligence, stability and interactivity;

[0047] Safety includes the functions, appearance and speed of the driverless vehicle; intelligence is reflected in the external environment, the judgment of traffic signal signs and driving tasks, the perception of the surrounding environment through Autoware's algorithm, the detection of moving vehicles and pedestrians, the avoidance of traffic accidents, and the detection of traffic lights and traffic police gesture recognition to avoid violating traffic rules; stability includes speed changes and steering return during driving; interactivity is used for communication between the vehicle and the user;

[0048] The safety scoring criteria include detecting whether the functions of the driverless vehicle are intact after a collision, whether the driverless vehicle is damaged after completing the driving mission, and whether the driverless vehicle drives at the prescribed speed under bad weather and road rules, and whether it provides users with a comfortable experience;

[0049] The intelligence of the unmanned vehicle is systematically scored through the fuzzy comprehensive evaluation method for the accuracy of signal recognition, traffic police gesture recognition, pedestrian and non-motor vehicle recognition, appropriateness of vehicle distance maintenance, rationality of route changes, and correct recognition and reasonable imitation of special lane giving. Special lane giving refers to making reasonable way for ambulances during emergency rescue missions. The sound of the ambulance is judged by the Autoware algorithm, and the behavior of the front and rear vehicles is detected by the video surveillance system, and reasonable imitation of lane giving is implemented. The stability is scored by whether the vehicle speed changes smoothly throughout the driving process and whether the steering response speed is reasonable. Finally, the interactivity is scored through the interactive response between the unmanned vehicle and the user.

[0050] The intelligence includes signal signs, pedestrian and non-motor vehicle recognition, vehicle distance keeping, traffic police gesture recognition, route changes, and special lane giving. Among them, route changes use aerospace technology to obtain satellite images, and then use intelligent algorithms to extract information such as road congestion to avoid poor road conditions.

[0051] In step S4, when the summary evaluation module uses AI to conduct a comprehensive evaluation of the drone-driven vehicle, the difficulty level is divided according to the different road scenarios in the road adaptation division module, and the evaluation is made based on the principles in the evaluation result evaluation module. The execution accuracy and accuracy of the unmanned vehicle in different difficulty environments are evaluated, and the adaptability range of the unmanned vehicle on different roads is determined. In addition, comprehensive calculations are used to analyze the shortcomings and disadvantages of the unmanned vehicle to provide data reference for improvement.

[0052] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and substitutions can be made without departing from the technical principles of the present invention. These improvements and substitutions should also be regarded as the scope of protection of the present invention.

Claims

1. Comprehensive quantitative evaluation test road system for unmanned vehicles, characterized by: It includes vehicle information collection module, road adaptation division module, evaluation result evaluation module and summary evaluation module; The vehicle information collection module is used to collect, record and transmit various types of data in real time or offline, and control path planning and execution behavior; The road adaptation division module is divided into scenarios of different difficulty composed of different environmental factors in suburban driving environment and urban driving environment, providing data for the vehicle's optimal adaptation range and accuracy evaluation under the scenario; The evaluation result evaluation module provides data for intelligent recognition and execution behavior evaluation during vehicle driving tasks based on the judgment principle; The summary evaluation module, the comprehensive road adaptation division module, and the test result evaluation module provide a comprehensive evaluation of vehicle performance and provide improvement data references for vehicle disadvantages.

2. The unmanned vehicle comprehensive quantitative evaluation test road system according to claim 1, characterized in that: The road adaptation division module mainly comprises dynamic factors, traffic signs, real-time weather and road conditions for different environmental factors in suburban driving environment and urban driving environment. By analyzing and modeling the road condition data, the complexity of the real-time road condition is obtained.

3. The comprehensive quantitative evaluation test method for unmanned vehicles according to claim 1 is characterized in that: The following steps are involved: S1, the vehicle information collection module uses a variety of sensors and computer equipment on the body of the unmanned vehicle to collect, record and transmit various types of data in real time or offline, as well as control path planning and execution behavior; S2, using the road adaptation classification module to classify the test scenarios of the unmanned vehicle. The module is divided into scenarios composed of different environmental elements under the suburban driving environment unit and the urban driving environment unit, and the road conditions for vehicle testing are divided and graded; S3, using the evaluation result evaluation module, based on the four sets of judgment principles, evaluates the accuracy of the vehicle's intelligent recognition and execution behavior during the driving task under different road conditions; S4, the summary evaluation module weights the overall detection of the unmanned vehicle according to the vehicle information collection module, the road adaptation division module, and the evaluation result evaluation module according to the sequence relationship analysis method, and uses the comprehensive evaluation method to comprehensively evaluate the unmanned vehicle to obtain the final evaluation.

4. The comprehensive quantitative evaluation test method for unmanned vehicles according to claim 3 is characterized in that: In step S1, a variety of sensors are used to identify traffic signs, vehicles, road conditions and real-time weather, and computer equipment is used to organize data, build maps, and control the path planning and execution behavior of the unmanned vehicle.

5. The comprehensive quantitative evaluation test method for unmanned vehicles according to claim 3 is characterized in that: In step S2, the suburban driving environment unit and the urban driving environment unit both include dynamic vehicles and pedestrians, traffic road signs, real-time weather, road conditions and other elements; Among them, dynamic vehicles and pedestrians refer to the number of vehicles and pedestrians, which is reflected in the degree of road congestion and the interaction of unmanned vehicles; real-time weather includes sunny, rainy, snowy, foggy and dusty days, which is explored through the fusion of lidar, camera and millimeter-wave radar; road conditions include the width of the road, the unevenness of the road surface and the construction status of the road surface.

6. The comprehensive quantitative evaluation test method for unmanned vehicles according to claim 3 is characterized in that: In step S3, the judgment principles include safety, intelligence, stability and interactivity; Safety includes the functions, appearance and speed of driverless vehicles; intelligence is reflected in the external environment, the judgment of traffic signal signs and driving tasks, and the perception of the surrounding environment through Autoware's algorithm; stability includes speed changes and steering return during driving; interactivity is used for communication between the vehicle and the user.

7. The comprehensive quantitative evaluation test method for unmanned vehicles according to claim 6 is characterized in that: The intelligence includes signal signs, pedestrian and non-motor vehicle recognition, vehicle distance keeping, traffic police gesture recognition, route changes, and special lane giving. Among them, route changes use aerospace technology to obtain satellite images, and then use intelligent algorithms to extract information such as road congestion to avoid poor road conditions.

8. The comprehensive quantitative evaluation test method for unmanned vehicles according to claim 3 is characterized in that: In step S4, when the summary evaluation module conducts a comprehensive evaluation on the drone driving vehicle, the difficulty level is divided according to the different road scenarios in the road adaptation division module, and the evaluation is made according to the principles in the evaluation result evaluation module to evaluate the execution accuracy and correctness of the unmanned vehicle in different difficulty environments.

Citation Information

Cited By

  • Human-computer interaction method and system of unmanned vehicle

    CN119283910A