Method and system for constructing automatic driving index evaluation model

Abstract

The invention belongs to the technical field of automatic driving, and particularly relates to a method and system for constructing an automatic driving index evaluation model, and the method comprises the following steps: 1, constructing an automatic driving index system which comprises a total index, three first-level indexes and six second-level indexes; 2, dividing indexes of the automatic driving index system into different types of indexes, constructing data normalization models suitable for the different types of indexes respectively, and performing data normalization processing; and 3, constructing a comprehensive evaluation model according to the automatic driving index system and the data normalization model, determining first-level and second-level index weights, and calculating an automatic driving index value through the comprehensive evaluation model. According to the method, the development condition of automatic driving is comprehensively evaluated through a systematic and scientific method, and the method is used for evaluating and comparing the automatic driving technology application level of cities and regions.

Description

Technical Field

[0001] The present invention belongs to the technical field of autonomous driving, and specifically relates to a method and system for constructing an autonomous driving index evaluation model. Background Art

[0002] In the process of industrialization of autonomous driving, due to differences in policies and degree of openness, the development progress varies from place to place, and there is an urgent need to build a set of dynamically developing evaluation tools.

[0003] In the process of application and industrialization of autonomous driving in the fields of taxis, buses, freight, distribution, etc., the industry lacks applicable and reasonable evaluation tools. The development of autonomous driving is limited to "yes" or "no" demonstration applications, and it is impossible to conduct a quantitative assessment of the "good" or "bad" application effect. Without comprehensive evaluation, it is impossible to effectively evaluate and compare the application level of autonomous driving technology in cities and regions.

[0004] This patent aims to build a complete autonomous driving evaluation system. By comprehensively considering factors such as policies and regulations, degree of openness, and application scale, it proposes an efficient evaluation tool to comprehensively evaluate the development of autonomous driving in a systematic and scientific manner. It solves the current problems in the evaluation of autonomous driving development, such as the lack of unified standards and uncertainty in evaluation content, and provides effective technical support. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for constructing an autonomous driving index evaluation model. By comprehensively considering factors such as policies and regulations, degree of openness, and application scale, an efficient evaluation tool is proposed to comprehensively evaluate the development of autonomous driving in a systematic and scientific manner, and to evaluate and compare the application level of autonomous driving technology in cities and regions.

[0006] The technical solution adopted by the present invention is as follows:

[0007] A method for constructing an autonomous driving index evaluation model comprises the following steps:

[0008] Step 1: Construct an autonomous driving index indicator system, which includes an overall index, three first-level indicators and six second-level indicators;

[0009] Step 2: Construct normalization methods applicable to different types of indicators;

[0010] The step 2 comprises the following steps:

[0011] Step 201: Divide the indicators of the autonomous driving index indicator system into different types of indicators;

[0012] Step 202: construct data normalization models suitable for different types of indicators;

[0013] Step 203: performing data normalization processing on different types of indicators respectively through different data normalization models;

[0014] Step 3: Construct a comprehensive evaluation method for the autonomous driving index;

[0015] The step 3 comprises the following steps:

[0016] Step 301: construct a comprehensive evaluation model based on the autonomous driving index indicator system constructed in step 1 and the data normalization model constructed in step 2;

[0017] Step 302: Determine the weights of various indicators of the first-level indicators and the second-level indicators;

[0018] Step 303: After the weights are determined, the autonomous driving index value is calculated through a comprehensive evaluation model.

[0019] Furthermore, the three first-level indicators in step 1 are policies and regulations, degree of openness and scale of application, respectively. The policies and regulations include pilot support, legislative support and policy completeness. The degree of openness includes open mileage and scenario coverage. The scale of application includes the number of autonomous driving vehicles.

[0020] The pilot support, legislative support, complete policies, open mileage, scenario coverage, and number of autonomous driving vehicles are all secondary indicators.

[0021] Furthermore, the statistical method for the pilot support is to take cities as the object, accumulate the number of autonomous driving pilot demonstration supports they have received, and refer to the pilot list officially announced by the licensing or pilot organizing department. If one item exists, 1 point will be calculated, and the total score will be accumulated, with a maximum score of 10 points;

[0022] The statistical method for legislative support is to accumulate the number of legislations at the city level or in the city's subordinate districts. One legislation is worth 1 point, and the total score is accumulated. The highest score is 5 points.

[0023] The statistical method for the policy completeness is to count the number of autonomous driving related policies issued by the city, divide the number into five levels: 0-3, 4-6, 7-9, 10-12, and 13 or above, and assign them 1, 2, 3, 4, and 5 points respectively;

[0024] The statistical method of the open road mileage is to count the open urban road mileage and the highway section mileage, and calculate the sum of the two mileages;

[0025] The logarithmic function of f(x) = a·log b(cx+1) is used for scoring;

[0026] Among them, a = 8.80; b = 1.60; c = 0.005; X is the logarithmic function of kilometers for scoring; X is the mileage and

[0027] The statistical method for the scenario coverage is: if there is an autonomous taxi scenario, +3 points; if there is an autonomous bus / coach scenario, +3 points; if there is an autonomous truck scenario, +3 points; if there is an unmanned delivery vehicle scenario, +1 point, and the total score is 10 points;

[0028] The statistical method for the number of autonomous driving vehicles is to count the number of L3 and above autonomous driving vehicles that are undergoing testing, demonstration applications, and operations up to the time point. The average production cost of unmanned delivery vehicles is only about half of that of other autonomous driving vehicles, so unmanned delivery vehicles are converted at 0.5 vehicles;

[0029] The logarithmic function f(x) = a·log b(cx+1) is used for scoring

[0030] Among them, a=10.50; b=2.00; c=0.0001; X is the number of vehicles.

[0031] Furthermore, the list of pilot projects officially announced by the awarding or pilot organizing departments includes the Ministry of Transport's intelligent transportation pilot application pilot, the Ministry of Industry and Information Technology's intelligent connected vehicle demonstration zone, the national-level vehicle networking pilot zone, the Ministry of Housing and Urban-Rural Development and the Ministry of Industry and Information Technology's dual-intelligence pilot, and the Ministry of Natural Resources' intelligent connected vehicle high-precision map application pilot.

[0032] Furthermore, the indicator types in step 2 include linear cumulative indicators, hierarchical cumulative indicators and logarithmic cumulative indicators. The pilot support and legislative support are linear cumulative indicators, the policy completeness and scenario coverage are hierarchical cumulative indicators, and the open mileage and number of autonomous driving vehicles are logarithmic cumulative indicators.

[0033] The data normalization model in step 2 includes a linear cumulative model, a hierarchical cumulative model and a logarithmic cumulative model;

[0034] The straight line accumulation model is used to perform data normalization processing on the straight line accumulation index;

[0035] The hierarchical accumulation model is used to perform data normalization processing on the hierarchical accumulation index;

[0036] The logarithmic accumulation model is used to perform data normalization processing on the logarithmic accumulation index.

[0037] Furthermore, in step 3, the weight of the regulations and policies is 0.2, the weight of the degree of openness is 0.4, the weight of the application scale is 0.4, and the sum of the weights of the secondary indicators under the same primary indicator is 1.

[0038] Furthermore, the weight of the pilot support is 0.5, the sum of the weights of the legislative support and policy completeness is 0.5, the weight of the open mileage is 0.75, the weight of the scenario coverage is 0.25, and the weight of the number of autonomous driving vehicles is 1.

[0039] Furthermore, the calculation formula of the autonomous driving index value in step 3 is as follows:

[0040]

[0041] Among them, V i and W i They are the first-level indicator data and the first-level indicator weight, v i and w i They are secondary indicator data and secondary weights respectively.

[0042] An autonomous driving index evaluation system, comprising:

[0043] The data collection module is used to collect data on the first-level and second-level indicators of autonomous driving in each city;

[0044] The indicator calculation module is used to calculate the value of each secondary indicator based on the collected data;

[0045] A comprehensive evaluation module, the comprehensive evaluation module comprising a linear accumulation calculation module, a hierarchical accumulation calculation module, a logarithmic accumulation calculation module and an autonomous driving index calculation module;

[0046] The linear accumulation calculation module is used to perform data normalization processing on pilot support and legislative support;

[0047] The hierarchical accumulation calculation module is used to perform data normalization processing on policy completeness and scenario coverage;

[0048] The logarithmic accumulation calculation module is used to perform data normalization processing on the open mileage and the number of autonomous driving vehicles;

[0049] The autonomous driving index calculation module is used to calculate the autonomous driving index value according to the data normalization result, the primary index weight, and the secondary index weight;

[0050] The result display module is used to display the autonomous driving index evaluation results of each city.

[0051] The technical effects achieved by the present invention are:

[0052] The method and system for constructing an autonomous driving index evaluation model of the present invention comprehensively considers factors such as policies and regulations, degree of openness, and application scale, and proposes an efficient evaluation tool to comprehensively evaluate the development of autonomous driving in a systematic and scientific manner, so as to evaluate and compare the application level of autonomous driving technology in cities and regions. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 is a flow chart of the steps of the present invention;

[0054] Figure 2 is a step flow chart of step 2 of the present invention;

[0055] Figure 3 is a step flow chart of step 3 of the present invention;

[0056] Figure 4 It is a schematic diagram of logarithmic function algorithm calculation of the present invention;

[0057] Figure 5 This is a graph showing the calculation results of Shanghai's autonomous driving index in the past three years in Example 2 of the present invention;

[0058] Figure 6 This is a graph showing the calculation results of Beijing’s autonomous driving index in the past three years in Example 2 of the present invention;

[0059] Figure 7 It is a system diagram of the present invention. DETAILED DESCRIPTION

[0060] In order to make the purpose and advantages of the present invention more clearly understood, the present invention is specifically described below in conjunction with embodiments. It should be understood that the following text is only used to describe one or several specific embodiments of the present invention, and does not strictly limit the scope of protection of the specific claims of the present invention.

[0061] In order to promote the healthy development of autonomous driving in the field of road traffic, scientifically measure and evaluate the development of L3 and above autonomous driving, and carry out research on the autonomous driving index, this patent mainly includes the following contents: construct an autonomous driving index indicator system. Construct a quantitative measurement model suitable for the dynamic development evaluation of autonomous driving, and realize the calculation of the autonomous driving index based on data processing methods such as linear accumulation, hierarchical accumulation, and logarithmic function.

[0062] The core content of this patent is to establish a multi-dimensional index indicator system applicable to autonomous driving policies, technologies and applications, and then construct an autonomous driving index evaluation method. This evaluation system provides a quantitative tool for transportation departments, enterprises, research institutions and the public to evaluate and compare the application level of autonomous driving technology in cities and regions. Through this research result, governments and enterprises can formulate targeted policies and strategies to further promote the innovative application of autonomous driving in cities and regions, and improve transportation service capabilities and public service levels.

[0063] This patent proposes a set of quantitative comprehensive evaluation models applicable to different indicators. This domestically pioneered evaluation method combines evaluation models of different indicators such as linear accumulation, graded accumulation, and logarithmic function, solving the problem of theoretical deficiency in existing evaluation methods. Through this set of comprehensive evaluation methods, scientific theoretical analysis tools can be provided for the dynamic development of autonomous driving, further promoting the innovative application and industrial development of autonomous driving technology in cities and regions.

[0064] Embodiment 1:

[0065] like Figure 1-3 As shown, a method for constructing an autonomous driving index evaluation model includes the following steps:

[0066] Step 1: Build an indicator system for the autonomous driving index.

[0067] Construct an autonomous driving index indicator system that covers multiple dimensions and factors such as policies and regulations, degree of openness, and scale of application; this patent clarifies the definition and connotation of each indicator, the statistical method of each indicator, the data content, etc. Finally, a complete autonomous driving index indicator system is constructed.

[0068] Specifically, the indicator system of the autonomous driving index in step 1 includes a total index, three first-level indicators and six second-level indicators. The three first-level indicators are policies and regulations, degree of openness and application scale, which respectively reflect the progress of autonomous driving policy support, open development and demonstration application; the six second-level indicators are pilot support, legislative support, policy completeness, open mileage, scenario coverage and number of autonomous driving vehicles, which are subdivided indicators of policies, degree of openness and demonstration application.

[0069] As shown in Table 1, in step 1, the focus is on the two dimensions of openness and application scale for evaluation, highlighting industry applications, the objectivity, accessibility, and accuracy of data, and promoting the development and application of autonomous driving technology.

[0070] Table 1: Autonomous Driving Index Indicator System

[0071]

[0072] Among them, policies and regulations include three secondary indicators: pilot support, legislative support, and complete policies.

[0073] 1① Pilot support: reflects the role of autonomous driving related demonstration pilot projects at the ministerial level.

[0074] Statistical method: Take cities as the objects, accumulate the number of autonomous driving pilot demonstrations they have obtained, and refer to the pilot list officially announced by the licensing or pilot organizing department. If one of the items exists, 1 point will be calculated, and the total score will be accumulated. The maximum score is 10 points.

[0075] Data content:

[0076] ·The Ministry of Transport’s pilot program for intelligent transportation applications;

[0077] · Intelligent connected vehicle demonstration zone and national-level vehicle networking pilot zone of the Ministry of Industry and Information Technology;

[0078] The dual-intelligence pilot projects of the Ministry of Housing and Urban-Rural Development and the Ministry of Industry and Information Technology;

[0079] The Ministry of Natural Resources' pilot project on high-precision map applications for intelligent connected vehicles;

[0080] · Pilot projects for the access and on-road use of intelligent connected vehicles by the Ministry of Industry and Information Technology, the Ministry of Public Security, the Ministry of Housing and Urban-Rural Development, and the Ministry of Transport;

[0081] The vehicle-road-cloud integration pilot project of the Ministry of Industry and Information Technology, the Ministry of Public Security, the Ministry of Natural Resources, the Ministry of Housing and Urban-Rural Development, and the Ministry of Transport

[0082] Data source: Pilot announcements on the official websites of the Ministry of Transport, the Ministry of Industry and Information Technology, the Ministry of Housing and Urban-Rural Development, the Ministry of Natural Resources, and the Ministry of Public Security

[0083] Data collection frequency and format: annual, document announcement

[0084] Data range: 0-10.

[0085] ② Legislative support: evaluate the legislative environment’s support for autonomous driving applications;

[0086] Statistical method: The number of laws at the city level or in the districts under the city. One piece of legislation is worth 1 point, and the total score is accumulated. The highest score is 5 points.

[0087] Data content: Legislation related to autonomous driving; including city / district-level legislation on intelligent connected vehicles, autonomous driving demonstration applications, etc.

[0088] Data range estimation: 0-5

[0089] Data collection frequency and format: annual, document announcement

[0090] Data source: Autonomous driving legislation and related policies of various cities are collected through government documents and the Internet.

[0091] ③ Policy completeness: Evaluate the policy environment’s support for autonomous driving applications

[0092] Statistical method: Policy quantity score: Count the number of autonomous driving related policies issued by the city, and divide the number into 5 levels: 0-3, 4-6, 7-9, 10-12, 13 and above, and assign them 1, 2, 3, 4, and 5 points respectively.

[0093] Data content: Autonomous driving related policies: including development planning, demonstration management, application management, industrial development and other policies on topics such as intelligent connected vehicles and autonomous driving.

[0094] Data range estimation: 0-5

[0095] Data collection frequency and format: annual, document announcement

[0096] Data source: Autonomous driving legislation and related policies of various cities are collected through government documents and the Internet.

[0097] Among them, the openness index includes two secondary indicators: open road mileage and scene coverage.

[0098] ① Open road mileage: The mileage of open roads in the city used to support autonomous driving test verification and demonstration applications. It reflects the city’s openness to autonomous driving.

[0099] Statistical method: Count the mileage of open urban roads; count the mileage of expressway sections, and calculate the sum of the two mileages.

[0100] The logarithmic function of f(x) = a·log b(cx+1) is used for scoring;

[0101] a=8.80; b=1.60; c=0.005; X is the logarithmic function of kilometers for scoring; X is the mileage.

[0102] Data content: Contains the mileage and length of urban roads and highways that support autonomous driving test verification and demonstration applications.

[0103] Data range estimation: 0-30 points

[0104] Data collection frequency and format: annual, kilometers

[0105] Data source: Local reports, Internet collection of open urban road mileage and open expressway mileage in each city

[0106] Calculation basis: Among the 14 cities, the total mileage of urban roads and highways is relatively concentrated, except for Guangzhou, which is the largest at 26,480 kilometers; it is about 6,000-9,000 kilometers. For this reason, 8,000 kilometers is set as the full score for this indicator.

[0107] ② Scenario coverage: The coverage of travel services, freight transportation, delivery services and other scenarios in the city’s autonomous driving demonstration applications or operations. This reflects the scope of autonomous driving application scenarios in different cities.

[0108] Statistical methods:

[0109] Self-driving taxi scenario (+3 points)

[0110] Self-driving bus / coach scenario (+3 points)

[0111] Self-driving truck scenario (+3 points)

[0112] Unmanned delivery vehicle scenario (+1 point). If there is one scenario, the corresponding score will be added, and the total score will be 10 points.

[0113] Data content: The coverage of travel services, freight transportation, delivery services and other scenarios in autonomous driving demonstration applications or operations. The higher the coverage, the higher the score, up to 10 points.

[0114] Data range estimation: 0-10 points

[0115] Data collection frequency and format: Annual, application report

[0116] Data source: Local reports and online collection of taxi, public transport / coach, truck, and unmanned delivery application scenarios in various cities.

[0117] Among them, application scale indicators include the number of autonomous driving vehicles.

[0118] ① Number of autonomous driving vehicles: The number of L3 and above autonomous driving vehicles that are being tested, demonstrated, and operated in the city. It represents the number and scale of L3 and above autonomous driving vehicles in different cities.

[0119] Statistical method: Statistics are based on the number of L3 and above autonomous driving vehicles that are being tested, demonstrated, and operated as of the time point. The average production cost of unmanned delivery vehicles is only about half of that of other autonomous driving vehicles, so unmanned delivery vehicles are converted at 0.5 vehicles.

[0120] The logarithmic function f(x) = a·log b(cx+1) is used for scoring

[0121] a=10.50;b=2.00;c=0.0001;X is the number of vehicles

[0122] Data content: Number of autonomous vehicles including: taxis, public transport / coaches, trucks, and unmanned delivery vehicles

[0123] Data coverage includes L3 and above autonomous driving vehicles that are undergoing testing, demonstration applications, and operations.

[0124] Data range estimation: 0-40 points

[0125] Data collection frequency and format: annual, number of vehicles

[0126] Data source: Local reports and online collection of the number of L3 and above self-driving taxis, buses, trucks, and unmanned delivery vehicles

[0127] Calculation basis: Based on the 2 million Tesla FSDs in the United States and the 240 million cars in the United States, the proportion of autonomous driving in China's 340 million market is about 2.4 million vehicles; if the growth rate in the next 15 years is 4%, it will be about 4.42 million vehicles. 442 / 13 = 130,000 vehicles / each administrative district.

[0128] During the measurement process, the measurement range is selected according to the following standards:

[0129] City selection criteria: The pilot cities related to autonomous driving will be mainly the first batch of intelligent transportation pilot application pilot cities of the Ministry of Transport, with some other cities as supplementary cities.

[0130] City selection principles:

[0131] ·The policy foundation for autonomous driving is relatively good

[0132] Demonstration application has achieved certain results

[0133] Larger cities

[0134] Step 2: Construct normalization methods applicable to different types of indicators.

[0135] Please refer to Figure 1 and Figure 2 As shown, in step 2, this technical solution combines the data value size and change characteristics of the secondary indicators constructed by the autonomous driving index indicator system, divides the autonomous driving indicator types, and combines different types to build their own applicable models for data normalization.

[0136] For example, the two indicators of policy completeness and scenario coverage are suitable for hierarchical accumulation; the indicators of open mileage, application scale, etc. in the autonomous driving index data are all cumulative values, and there is no maximum value in theory. The magnitude of such data far exceeds that of other indicators. If the traditional normalized de-dimensionalization calculation method is used for data processing, the characteristics of such data will be destroyed, which is not conducive to promoting the completeness and scientificity of the final autonomous driving index. Therefore, logarithmic accumulation is studied to solve the normalization problem of such indicators. Finally, the research forms a data normalization processing method applicable to different types of indicators.

[0137] Among them, the level of autonomous driving in most cities in my country is still at a relatively early stage, and it is a critical breakthrough period for autonomous driving to achieve a transition from 0 to 1. The autonomous driving index should encourage more cities to develop autonomous driving technology, rather than allowing developed cities to always be far ahead.

[0138] Different types of indicators should adopt different calculation methods, and appropriate algorithm models should be developed based on the data characteristics of different types of indicators.

[0139] Among them, the two indicator data of pilot support and legislative support in the autonomous driving index data are both cumulative values. Theoretically, with the continuous development of my country's autonomous driving technology, the number of pilot support and legislative support will continue to increase, and there is no upper limit to the indicator score. Therefore, pilot support and legislative support are classified as straight-line cumulative indicators, and a straight-line cumulative model is constructed to normalize such indicator data. Based on the comprehensive consideration of the basic national conditions that the level of autonomous driving in most cities in my country is still in a relatively early stage, the corresponding theoretical maximum value is set to facilitate subsequent summary calculations. The relevant scoring method of the straight-line cumulative model is shown in the following table:

[0140] Table 2 Linear cumulative index score table

[0141]

[0142]

[0143] Among them, the two indicators of policy completeness and scenario coverage in the autonomous driving index data are both graded cumulative values. The two indicators of policy completeness and scenario coverage are classified as graded cumulative indicators, and a graded cumulative model is constructed to normalize such indicator data. These two indicators divide the relevant data into several levels or categories, and each level (category) is scored separately. The relevant scoring method of the graded cumulative indicators is shown in the following table:

[0144] Table 3 Grading cumulative index score table

[0145]

[0146] Among them, the open mileage, application scale and other indicators in the autonomous driving index data are all cumulative values, and there is no maximum value in theory. The magnitude of such data far exceeds that of other indicators. If the traditional normalized de-dimensionalization calculation method is used for data processing, the characteristics of such data will be destroyed, which is not conducive to promoting the completeness and scientificity of the final autonomous driving index.

[0147] Taking into account the basic national conditions that the level of autonomous driving in most cities in my country is still in a relatively early stage, and it is also a critical breakthrough period for most cities to achieve a breakthrough from 0 to 1; to ensure that the index score of the cumulative value can achieve a monotonically increasing trend, and considering the basic conditions of most cities, more cities are encouraged to develop autonomous driving technology, and finally the open mileage and application scale are classified into logarithmic cumulative indicators, and a logarithmic cumulative model is constructed to normalize such indicator data. The logarithmic cumulative model uses a logarithmic function algorithm for scoring. Specifically, Figure 4 shown.

[0148] ① Dimensionless processing of open road mileage indicators

[0149] Using the formula:

[0150] X norm =a log b (cX+1)

[0151] The logarithmic function of is used to score, and the parameters a, b, and c are fitted based on the overall situation of the indicator data; X is the number of kilometers; where X is the sum of the mileage of open urban roads and open highways.

[0152] ②Dimensionless processing of the number of autonomous driving vehicles

[0153] Using the formula:

[0154] X norm =a log b (cX+1)

[0155] The logarithmic function is used to score, and the parameters a, b, and c are fitted based on the overall situation of the indicator data; X is the number of vehicles; where X is the number of autonomous driving vehicles and. It includes the number of four types of vehicles, namely, L3 and above autonomous driving taxis, public buses, trucks, and unmanned delivery vehicles, which are undergoing testing, demonstration applications, and operations. The average production cost of unmanned delivery vehicles is only about half of that of other autonomous driving vehicles, so unmanned delivery vehicles are converted at 0.5.

[0156] Step 3: Construct a comprehensive evaluation method for the autonomous driving index;

[0157] Please refer to Figure 1 and Figure 3 As shown, step 3 includes the following steps:

[0158] Step 301: constructing a comprehensive evaluation model including a linear accumulation model, a hierarchical accumulation model, a logarithmic accumulation model, etc. according to the autonomous driving index indicator system constructed in step 1 and the data normalization model constructed in step 2;

[0159] Step 302: Determine the weights of various indicators of the first-level indicators and the second-level indicators;

[0160] Among them, in step 3, the weights of the secondary indicators are fixed; the primary indicators are set to 0.2, 0.4, and 0.4 according to expert opinions, and the sum of the weights of the secondary indicators under the same primary indicator is 1.

[0161] Table 4 Index weight table

[0162]

[0163] Step 303: After the weights are determined, the automatic driving index value is calculated through the comprehensive evaluation model;

[0164] The method for calculating the value of the autonomous driving index in step 3 includes the following steps:

[0165] The autonomous driving index is calculated based on the weights of the first- and second-level indicators in the data set. The weights of the first-level indicators are set to 0.2, 0.4, and 0.4, and the weights of the second-level indicators are set to the sum of the weights of the second-level indicators under the same first-level indicator is 1.

[0166] The calculation formula is as follows:

[0167]

[0168] Among them, V i and W i They are the first-level indicator data and the first-level indicator weight, v i and w i They are secondary indicator data and secondary weights respectively.

[0169] The autonomous driving index in this embodiment focuses on the development and application of autonomous driving in different cities. It measures that different cities have basically the same autonomous driving technology paths, but have different progress in autonomous driving applications due to different policies and open driving foundations. It incorporates quantitative indicators of application scale and adopts a subjective + objective comprehensive evaluation.

[0170] Embodiment 2:

[0171] This embodiment discloses a process for calculating an autonomous driving index based on Embodiment 1, which is as follows:

[0172] The collected raw data are as follows:

[0173] Table 5. Raw data of cities

[0174]

[0175]

[0176] Automated Driving Index Calculation:

[0177] Table 6 Original values ​​and weights of secondary indicators

[0178]

[0179] Table 7 Calculation results and weights of primary indicators

[0180]

[0181] According to the first-level indicator weights of 0.2, 0.4, and 0.4, the calculation results show that the highest score of the autonomous driving index is 40.17 points in Shanghai, and the lowest score is 12.38 points in Xiamen.

[0182] The top five cities in autonomous driving are: Shanghai, Wuhan, Guangzhou, Beijing, and Suzhou. A total of 14 cities were surveyed, and the rankings from high to low according to scores are: Shanghai, Wuhan, Guangzhou, Beijing, Suzhou, Shenzhen, Tianjin, Chongqing, Hefei, Wuxi, Changsha, Hangzhou, Xi'an, and Xiamen.

[0183] Table 8 Automated Driving Index Results

[0184]

[0185]

[0186] In order to verify the rationality and scientificity of the autonomous driving index system, the development of autonomous driving in Shanghai and Beijing in the past three years was investigated and calculated. Figure 5-6 As shown;

[0187] From the calculation results, it can be seen that the autonomous driving index of Shanghai and Beijing has shown a stable increasing trend, and the values ​​of indicators at all levels have not shown unreasonable declines or step-like phenomena. The calculation process of the autonomous driving index is effective and reasonable.

[0188] In this embodiment, the autonomous driving index ranges from 0 to 100 points. It is necessary to set quantitative benchmarks of 30 points, 60 points, and 100 points based on the actual development of autonomous driving, to assist in comparing and explaining the differences or similarities between different cities, and to more clearly show the performance of autonomous driving development.

[0189] ① The current average level of autonomous driving = 30 points. At present, my country's special policies for autonomous driving have been implemented first, basically meeting the development needs. The city is open in a small regional area, and the mileage of open roads accounts for about 10% of the total mileage. The scene application is not sufficient, the application scale is small, and there are about 200-300 L3 and above autonomous driving vehicles. The current average level of autonomous driving is set to 30 points. Shanghai, Beijing, Guangzhou, Wuhan and other cities belong to the first echelon, which is higher than the average, about 30-40 points; Tianjin, Shenzhen, Hefei, Chongqing, Suzhou, Hangzhou are in the second echelon, with a score of about 20-30 points; Changsha, Xi'an, Xiamen are in the third echelon, with a score of less than 20 points.

[0190] ② The autonomous driving standard line = 60 points. The autonomous driving development standard line is set at 60 points, and the corresponding indicators of 60 points are: sufficient support for autonomous driving special policies, large-scale opening of cities, open road mileage accounting for more than 50% of the total mileage, comprehensive application of 5 types of scenarios, initial application, and more than 10,000 L3 and above autonomous driving vehicles. It is expected that first-tier cities such as Shanghai and Beijing will reach this level in 3-5 years.

[0191] ③ Full score for autonomous driving development = 100 points. It is expected that in 15 years, the special policies for autonomous driving will be comprehensive and sufficient, cities will be fully open to autonomous driving, 5 types of scenarios will be fully applied, and there will be more than 130,000 L3 and above autonomous driving vehicles.

[0192] Table 9 Autonomous Driving Index Benchmark Setting and Prediction Table

[0193]

[0194]

[0195] Analysis of the Autonomous Driving Index results:

[0196] The indicator score ranges are different, and the application scale is the most important. From the perspective of the score range of the first-level indicators, the application scale and degree of openness indicator ranges are twice the policy and regulatory indicator ranges. The score ranges of the three first-level indicators of policies and regulations, degree of openness, and application scale are 0-20 points, 0-40 points, and 0-40 points, respectively. From the perspective of the second-level indicators, the score ranges of the six second-level indicators of pilot support, legislative support, policy completeness, open mileage, scenario coverage, and number of autonomous driving vehicles are: 0-10 points, 0-5 points, 0-5 points, 0-20 points, 0-20 points, and 0-40 points, respectively. The interval range of the second-level indicator number of autonomous driving vehicles is 8 times the interval of the legislative support / policy completeness indicators and 4 times the interval range of the pilot support indicator.

[0197] Table 10 Automated Driving Index Range Results

[0198]

[0199] The autonomous driving index is generally positively correlated with policies, openness, and application scale. Cities such as Shanghai, Wuhan, Guangzhou, and Beijing rank high, thanks to their comprehensive strength in autonomous driving policy support, openness, and the scale of autonomous driving applications such as taxis. Changsha, Hangzhou, Xi'an, and Xiamen rank low, mainly due to the large gap in the scale of their autonomous driving applications, and their unclear policies and openness. Changsha has only more than 40 autonomous driving taxis in demonstration applications, while the total number of autonomous driving vehicles for taxis, public buses, and trucks in Xi'an and Xiamen is only in the single digit, and the scale of application is seriously small; Hangzhou has a small mileage of open roads, only 100+ kilometers, which leads to the index ranking being low.

[0200] Overall, there is still a lot of room for improvement in the scale of autonomous driving applications. At present, the highest scores of the three first-level indicators of policy, degree of openness, and application scale are approximately 17 points, 28 points, and 1 point, respectively. According to the range of first-level indicators, the highest score of the application scale should be equal to the degree of openness, which is twice the policy score. The above shows that based on the current urban autonomous driving policy, the application scale indicator of the number of autonomous driving vehicles should theoretically be higher. At present, with the support of good policy drive and open environment, the demonstration application of autonomous driving is insufficient. The effective improvement of autonomous driving demonstration applications requires focusing on the innovation of autonomous driving business models, the drive of autonomous driving industry, and the management of autonomous driving applications, rather than simply autonomous driving pilots, legislation, policy drive and road opening.

[0201] Embodiment 3:

[0202] See also Figure 7 As shown, this embodiment discloses an autonomous driving index evaluation system based on the above embodiment, and the autonomous driving index evaluation system includes a data collection module, an index calculation module, a comprehensive evaluation module and a result display module;

[0203] Among them, the data collection module is used to connect with the databases of relevant departments such as the Ministry of Transport, the Ministry of Industry and Information Technology, the Ministry of Housing and Urban-Rural Development, and the Ministry of Natural Resources to obtain information such as the officially announced pilot list, and collect relevant data on the first-level and second-level indicators of autonomous driving in each city, including pilot demonstration support, number of legislation, number of policies, open road mileage, scenario coverage, and the number of autonomous driving vehicles.

[0204] Among them, the indicator calculation module is used to calculate the value of each secondary indicator based on the collected data.

[0205] For pilot support, cities are taken as the target, and the number of autonomous driving pilot demonstration supports they receive is accumulated, and points are calculated based on the pilot list officially announced by the awarding or pilot organizing department.

[0206] Legislative support is scored by accumulating the number of pieces of legislation at the city level or in the districts within the city.

[0207] Policy completeness is graded and assigned values ​​based on the number of autonomous driving-related policies issued.

[0208] The open mileage is calculated using a specific logarithmic function, which counts the sum of the open urban road mileage and highway section mileage.

[0209] The scene coverage is scored cumulatively based on the existence of different scenes.

[0210] The number of autonomous driving vehicles includes the number of L3 and above autonomous driving vehicles. Unmanned delivery vehicles are converted to 0.5 vehicles and are also scored using a specific logarithmic function.

[0211] Among them, the comprehensive evaluation module includes a linear accumulation calculation module, a hierarchical accumulation calculation module, a logarithmic accumulation calculation module and an autonomous driving index calculation module;

[0212] The linear accumulation calculation module is used to normalize the data of pilot support and legislative support;

[0213] The hierarchical accumulation calculation module is used to normalize the data of policy completeness and scenario coverage;

[0214] The logarithmic accumulation calculation module is used to normalize the data of open mileage and the number of autonomous driving vehicles;

[0215] The autonomous driving index calculation module is used to calculate the autonomous driving index value based on the data normalization results, the first-level indicator weights, and the second-level indicator weights.

[0216] Among them, the result display module is used to display the autonomous driving index evaluation results of each city in the form of intuitive charts, including the total index, first-level indicator scores and second-level indicator scores, so that users can understand the advantages and disadvantages of different cities in the development of autonomous driving.

[0217] In summary, this technical solution proposed the construction of an autonomous driving index for the first time in China, providing an innovative multi-dimensional comprehensive indicator system for dynamically evaluating the development of autonomous driving technology. This system is comprehensive and highly integrated, filling the gap in this field in China. The research focus of this project is to build a complete set of autonomous driving evaluation tools, covering the autonomous driving index indicator system and index evaluation method model, aiming to solve the comprehensive development evaluation problems faced by autonomous driving in the application and industrialization of taxis, buses, freight and distribution. The research not only constructed this indicator system, but also originally proposed an autonomous driving index evaluation method, and formed an effective comprehensive evaluation tool by classifying and constructing data processing methods such as linear accumulation, hierarchical accumulation, and logarithmic functions. This research is not only technologically innovative and economically valuable, but can also promote the development of a trillion-level autonomous driving industry including vehicles, chips, and roadside equipment. The innovative results of the research include the following:

[0218] 1) Autonomous Driving Index System: This is the first multi-dimensional index system in China that can evaluate the dynamic development of autonomous driving technology in different cities and regions. Due to the differences in policies and openness among regions, this index system can accurately reflect the progress of different regions and solve the problems of missing indicators and uncertain evaluation content in the current evaluation.

[0219] 2) Autonomous driving index evaluation method: This patent constructs for the first time a variety of data processing methods suitable for autonomous driving evaluation, such as linear accumulation, hierarchical accumulation and logarithmic function, forming a comprehensive evaluation model, which makes up for the lack of existing evaluation method theory. This method provides theoretical support for the innovative application of autonomous driving technology in urban and regional fields.

[0220] This patent not only provides important support for the theoretical innovation and application exploration of the National Key Laboratory in the field of autonomous driving, but also lays the foundation for building a Chinese paradigm. Through close connection with industry, academia and research, this patent is expected to accelerate the research and development of key technologies and the transformation of results, and further enhance my country's independent innovation capabilities in this field.

[0221] This patent has important application value in many aspects, and can promote social and economic development, enhance government governance capabilities, promote industry technological progress, and help industrial development. The main application prospects are as follows:

[0222] 1) National autonomous driving development evaluation: This patent can provide a quantitative tool for the development of autonomous driving across the country, helping governments and companies to formulate targeted policies, promote the innovative application of autonomous driving in cities and regions, and enhance transportation service capabilities.

[0223] 2) Evaluation of the development of autonomous driving in urban areas: This patent can help analyze the key points and shortcomings of the development of autonomous driving in urban areas, deeply evaluate the gaps in policies and regulations, degree of openness, application scale, etc., and formulate targeted strategies to promote the demonstration application of autonomous driving technology in scenarios such as taxis, buses, and logistics.

[0224] 3) Comparative analysis of autonomous driving at home and abroad: Through the study of the autonomous driving index, we can support comparative analysis between domestic and foreign cities, provide technical support on policies, degree of openness and application, and promote the international development of autonomous driving technology.

[0225] The above is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications should also be considered as the protection scope of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention shall be implemented according to the conventional means in the art unless otherwise specified and limited.

Claims

1. A method for constructing an autonomous driving index evaluation model, characterized in that: The following steps are involved: Step 1: Construct an autonomous driving index indicator system, which includes an overall index, three first-level indicators and six second-level indicators; Step 2: Construct normalization methods applicable to different types of indicators; The step 2 comprises the following steps: Step 201: Divide the indicators of the autonomous driving index indicator system into different types of indicators; Step 202: construct data normalization models suitable for different types of indicators; Step 203: performing data normalization processing on different types of indicators respectively through different data normalization models; Step 3: Construct a comprehensive evaluation method for the autonomous driving index; The step 3 comprises the following steps: Step 301: construct a comprehensive evaluation model based on the autonomous driving index indicator system constructed in step 1 and the data normalization model constructed in step 2; Step 302: Determine the weights of the primary index and the secondary index; Step 303: After the weights are determined, the autonomous driving index value is calculated through a comprehensive evaluation model.

2. The method for constructing an autonomous driving index evaluation model according to claim 1, characterized in that: The three first-level indicators in step 1 are policies and regulations, degree of openness and application scale. The policies and regulations include pilot support, legislative support and policy completeness. The degree of openness includes open mileage and scenario coverage. The application scale includes the number of autonomous driving vehicles. The pilot support, legislative support, complete policies, open mileage, scenario coverage, and number of autonomous driving vehicles are all secondary indicators.

3. The method for constructing an autonomous driving index evaluation model according to claim 2, characterized in that: The statistical method for the pilot support is to take cities as the object and accumulate the number of autonomous driving pilot demonstration supports they have received, based on the pilot list officially announced by the licensing or pilot organizing department. If one item exists, 1 point will be calculated, and the total score will be accumulated, with a maximum score of 10 points; The statistical method for legislative support is to accumulate the number of legislations at the city level or in the city's subordinate districts. One legislation is worth 1 point, and the total score is accumulated. The highest score is 5 points. The statistical method for the policy completeness is to count the number of autonomous driving related policies issued by the city, divide the number into five levels: 0-3, 4-6, 7-9, 10-12, and 13 or above, and assign them 1, 2, 3, 4, and 5 points respectively; The statistical method of the open road mileage is to count the open urban road mileage and the highway section mileage, and calculate the sum of the two mileages; The logarithmic function of f(x) = a·log b(cx+1) is used for scoring; Among them, a = 8.80; b = 1.60; c = 0.005; X is the logarithmic function of kilometers for scoring; X is the mileage; The statistical method for the scenario coverage is: if there is an autonomous taxi scenario, +3 points; if there is an autonomous bus / coach scenario, +3 points; if there is an autonomous truck scenario, +3 points; if there is an unmanned delivery vehicle scenario, +1 point, and the total score is 10 points; The statistical method for the number of autonomous driving vehicles is to count the number of L3 and above autonomous driving vehicles that are undergoing testing, demonstration applications, and operations up to the time point. The average production cost of unmanned delivery vehicles is only about half of that of other autonomous driving vehicles, so unmanned delivery vehicles are converted at 0.5 vehicles; The logarithmic function f(x) = a·log b(cx+1) is used for scoring Among them, a=10.50; b=2.00; c=0.0001; X is the number of vehicles.

4. The method for constructing an autonomous driving index evaluation model according to claim 2, characterized in that: The list of pilot projects officially announced by the awarding or pilot organizing departments includes the Ministry of Transport's intelligent transportation pilot application pilot, the Ministry of Industry and Information Technology's intelligent connected vehicle demonstration zone, the national-level vehicle networking pilot zone, the Ministry of Housing and Urban-Rural Development and the Ministry of Industry and Information Technology's dual-intelligence pilot, and the Ministry of Natural Resources' intelligent connected vehicle high-precision map application pilot.

5. The method for constructing an autonomous driving index evaluation model according to claim 2, characterized in that: The indicator types in step 2 include linear cumulative indicators, hierarchical cumulative indicators and logarithmic cumulative indicators. The pilot support and legislative support are linear cumulative indicators, the policy completeness and scenario coverage are hierarchical cumulative indicators, and the open mileage and number of autonomous driving vehicles are logarithmic cumulative indicators. The data normalization model in step 2 includes a linear cumulative model, a hierarchical cumulative model and a logarithmic cumulative model; The straight line accumulation model is used to perform data normalization processing on the straight line accumulation index; The hierarchical accumulation model is used to perform data normalization processing on the hierarchical accumulation indicators; The logarithmic accumulation model is used to perform data normalization processing on the logarithmic accumulation index.

6. The method for constructing an autonomous driving index evaluation model according to claim 2, characterized in that: In step 3, the weight of the regulations and policies is 0.2, the weight of the degree of openness is 0.4, the weight of the application scale is 0.4, and the sum of the weights of the secondary indicators under the same primary indicator is 1.

7. The method for constructing an autonomous driving index evaluation model according to claim 6, characterized in that: The weight of the pilot support is 0.5, the sum of the weights of legislative support and policy completeness is 0.5, the weight of the open mileage is 0.75, the weight of the scenario coverage is 0.25, and the weight of the number of autonomous driving vehicles is 1.

8. The method for constructing an autonomous driving index evaluation model according to claim 6, characterized in that: The calculation formula of the autonomous driving index value in step 3 is as follows: Among them, V i and W i They are the first-level indicator data and the first-level indicator weight, v i and w i They are secondary indicator data and secondary weights respectively.

9. An autonomous driving index evaluation system, using a method for constructing an autonomous driving index evaluation model according to any one of claims 1 to 8, characterized in that: The autonomous driving index evaluation system includes: The data collection module is used to collect data on the first-level and second-level indicators of autonomous driving in each city; The indicator calculation module is used to calculate the value of each secondary indicator based on the collected data; A comprehensive evaluation module, the comprehensive evaluation module comprising a linear accumulation calculation module, a hierarchical accumulation calculation module, a logarithmic accumulation calculation module and an autonomous driving index calculation module; The linear accumulation calculation module is used to perform data normalization processing on pilot support and legislative support; The hierarchical accumulation calculation module is used to perform data normalization processing on policy completeness and scenario coverage; The logarithmic accumulation calculation module is used to perform data normalization processing on the open mileage and the number of autonomous driving vehicles; The autonomous driving index calculation module is used to calculate the autonomous driving index value according to the data normalization result, the primary index weight, and the secondary index weight; The result display module is used to display the autonomous driving index evaluation results of each city.

Images