Elderly driver suitable driving evaluation system construction method fusing three-force test

Abstract

The invention relates to a three-force test fused old-age driver suitable driving evaluation system construction method, which comprises the steps of obtaining questionnaire survey data of old-age drivers of more than 60 years old, performing descriptive statistical analysis and Pearson correlation analysis on the obtained data, and screening to obtain a high-risk driving scene; constructing a simulated driving scene, and searching suitable individuals to carry out simulated driving experiments and function tests to obtain experimental data; an evaluation index set is obtained based on simulated driving experiment data arrangement, and an old people driving suitability evaluation system is constructed from the perspective of three-force testing of memory, judgment and reactivity. The invention makes up for the defects of the existing three-force test question bank, provides a set of three-force test integrated old people driving suitability evaluation system and a new version of three-force test question bank and evaluation system integrated with dynamic risk driving scenes, effectively realizes three-force test for old people drivers, assists in screening high-risk old people drivers, and improves the driving suitability of the old people. The road safety risk of old drivers is reduced, and the practical application value is achieved.

Description

Technical Field

[0001] The present invention relates to the field of traffic safety technology, and in particular to a method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests. Background Art

[0002] The removal of the 70-year-old age limit for applying for a small car driver's license has led to an increase in the number of elderly drivers. However, advanced age can lead to a decline in physical function and a decline in attention and reaction ability. The road traffic safety problems that may be caused by this should not be underestimated.

[0003] In response to the issue of aging drivers, people over 70 years old need to pass the "three strengths" test of memory, judgment, and reaction before they can apply for a motor vehicle driver's license. However, domestic research on the fitness assessment of elderly drivers is still incomplete. The existing "three strengths" test questions are far from sufficient, especially unable to effectively assess the elderly's reaction ability and effectively identify high-risk elderly drivers. The road safety hazards of elderly drivers need to be addressed urgently.

[0004] Therefore, at the current stage, how to provide an effective elderly driver suitability assessment system and evaluation system based on the "three forces" test to screen high-risk elderly drivers and those who are unsuitable to drive is crucial to promoting the high-quality development of elderly travel safety and road traffic safety. Summary of the invention

[0005] The purpose of the present invention is to provide a method for evaluating the driving suitability of elderly drivers. The high-risk driving scenarios of the elderly are obtained through questionnaire surveys, and then relevant test data are obtained through simulated driving tests and functional tests, and a driving suitability evaluation system for the elderly is constructed from the perspective of three-force tests. The present invention makes up for the deficiency that most existing tests are too simple in terms of text description questions, and conducts multi-faceted ability tests on the memory, judgment, and reaction of elderly drivers, effectively screening high-risk elderly drivers and people who are not suitable for driving.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] The method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests described in the present invention comprises:

[0008] Step S1: Obtain survey data of elderly drivers over 60 years old through a questionnaire, perform descriptive statistical analysis and Pearson correlation analysis on the obtained data, and screen out high-risk driving scenarios;

[0009] Step S2: constructing a simulated driving scenario based on the questionnaire survey data analysis results, finding suitable individuals to conduct simulated driving experiments and functional tests, and obtaining simulated driving experiment data;

[0010] Step S3: Based on the simulated driving experiment data, an evaluation index set is obtained, and a driving suitability evaluation system for the elderly is constructed from the perspectives of memory, judgment, and reaction.

[0011] The method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests of the present invention comprises obtaining questionnaire survey data of elderly drivers over 60 years old, performing descriptive statistical analysis and Pearson correlation analysis on the obtained data, and screening out high-risk driving scenarios, including:

[0012] Step S1-1: Design a questionnaire on driving willingness and driving risk of the elderly based on the SP and RP survey methods, distribute the questionnaire online and offline at the same time and collect valid samples;

[0013] Step S1-2: Perform descriptive statistical analysis on the valid sample data, and use Excel and SPSS to perform data statistics; obtain the driving scenarios that most elderly drivers choose to avoid and the driving scenarios with risk self-assessment scores higher than the average as the high-risk driving scenario set 1;

[0014] Step S1-3: extract the avoidance choices of elderly drivers for different risk scenarios from the questionnaire, assign weighted scores to the questions and obtain the corresponding total driving risk scores, perform Pearson correlation analysis on the two, and select scenarios with correlation coefficients higher than 0.2 as high-risk driving scenario set 2;

[0015] Step S1-4: Based on the high-risk driving scenario set 1 obtained in step S1-2, the high-risk driving scenario set 2 obtained in step S1-3 is used as a subsequent simulation test scenario.

[0016] The method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests described in the present invention, constructing a simulated driving scenario based on the questionnaire survey data analysis results, finding suitable individuals for simulated driving experiments and functional tests, and obtaining experimental data, includes:

[0017] Step S2-1: construct four typical driving scenarios, and set multiple special events in each driving scenario; the four driving scenarios include: urban road section, highway, intersection, and rural road section;

[0018] Special events include: noise interference, heavy traffic flow, driving in rainy weather, mixed traffic of motor vehicles and non-motor vehicles, pedestrians rushing out without warning, turning at intersections without signal control, and stop and yield signs ahead;

[0019] Step S2-2: Find a suitable individual to conduct a simulated driving experiment and obtain simulated driving experiment data;

[0020] Step S2-3: guiding the person under test to perform a functional test and obtain functional test data;

[0021] Functional tests include: vision test, physical fitness test, and cognitive ability test.

[0022] The method for constructing a driving suitability evaluation system for elderly drivers integrating three-force tests described in the present invention is to obtain an evaluation index set based on simulated driving experimental data and divide it into three categories: memory, judgment, and reaction, so as to construct a driving suitability evaluation system for elderly drivers, including:

[0023] Based on high-risk driving scenario set 1 and high-risk driving scenario set 2, simulated driving scenarios were built on the computer using software and built into the laboratory's simulated driving device for the tester's on-site simulated driving experiment;

[0024] Step S3-1: The simulated driving test data of the test subjects are given to experts for scoring and screening, and the indicators with obvious differences between the unfit driving group and the fit driving group are screened out as the preliminary indicator set;

[0025] Step S3-2: Screen out indicators with a high correlation with the results of unsuitable driving samples from the remaining data information covered by the questionnaire, experiment and functional test through correlation analysis. The indicators are a secondary indicator set obtained through Pearson correlation analysis, including: functional stretch test, traffic accident last year, and self-assessment of driving ability. These three indicators are used as secondary indicator expansion to determine the evaluation indicator set;

[0026] The remaining data information refers to the previous questionnaire data, simulated driving experiment data, functional test data, etc. For example, the "whether there were traffic accidents last year" obtained through the questionnaire, the "maximum vehicle speed" and "steering wheel angle" obtained through the simulated driving test; Pearson correlation analysis is performed with the preliminary indicator set obtained from S3-1, and indicators with correlation coefficients higher than 0.5 are screened out as the secondary indicator set.

[0027] Step S3-3: taking the preliminary index set in step S3-1 and the evaluation index set expanded in step S3-2 as the final evaluation index; dividing the final evaluation index into four categories: memory, judgment, reaction, and comprehensive, and constructing an evaluation system for driving suitability of the elderly;

[0028] The functional stretch test index is derived from the data obtained from the functional test, and the two indicators of whether there was a traffic accident last year and self-assessment of driving ability are derived from the data obtained from the questionnaire survey.

[0029] Step S3-4: According to the obtained evaluation system, the data obtained from the questionnaire are applied to four models: factor analysis, entropy weight-TOPSIS method, CRITIC method, and neural network comprehensive evaluation. The data are compared with the expert scoring results to verify the rationality of the evaluation system from the perspective of binary classification and score ranking.

[0030] In the method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests described in the present invention, in step S3-3,

[0031] Evaluation indicators focus on memory including: eight-word test;

[0032] Judgment includes: observing traffic signs on unfamiliar roads, difficulty dealing with heavy traffic, difficulty dealing with rainy days, blind spot inspections, changes in the vehicle's position in the lane of travel, contact with other road users, speeding or failure to obey traffic lights;

[0033] Responsiveness includes: eye and head movements, cross-scanning;

[0034] Comprehensive includes: functional stretch test, traffic accidents in the past year, and self-assessment of driving ability.

[0035] In the method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests described in the present invention, in step S3-4, the comprehensive evaluation model expression of the factor analysis method is:

[0036]

[0037] Among them, ω i is the variance contribution rate of the ith factor, F 1j 、F 2j 、F 3j 、F 4j are the four factors whose characteristic roots of the j-th group of experimental data are greater than 1, F j is the comprehensive evaluation score of the jth experimenter.

[0038] In the method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests described in the present invention, in step S3-4, the steps for establishing an entropy weight-TOPSIS method comprehensive evaluation model are as follows:

[0039] Create the initial matrix:

[0040]

[0041] Among them, (X ij ) m×n are the n evaluation index values ​​of m elderly driver experimenters, i is the experimenter data label, and j is the evaluation index label;

[0042] Perform min-max normalization on the data in the evaluation matrix and convert negative indicators into positive indicators to obtain the normalized matrix:

[0043]

[0044] Among them, (Y ij )m×n is the normalized value of n evaluation indicators of m experimenters;

[0045] Calculate the information entropy and weight of each indicator, and weight the experimental data to obtain the weighted normalized matrix:

[0046]

[0047] Among them, (V ij ) m×n is the weighted and normalized value of n evaluation indicators of m experimenters;

[0048] Calculate the Euclidean distance from the evaluation object to the positive and negative ideal solutions, and calculate the relative progress:

[0049]

[0050] in, is the distance between the ith solution and the optimal solution, is the distance between the ith solution and the worst solution, C i is the relative progress of the i-th experimenter;

[0051] The experimental data were input into the entropy weight-TOPSIS comprehensive evaluation model to calculate the score of each experimenter. The dividing line for determining whether the elderly driver is suitable for driving is in the range of [0.514481, 0.520705].

[0052] In the method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests described in the present invention, in step S3-4, the steps for establishing the CRITIC method comprehensive evaluation model are as follows:

[0053] The experimental data was dimensionless processed, the standard deviation and conflict of each evaluation index were calculated, and the information amount calculation formula of each index was obtained:

[0054] C j =S j ×R j

[0055] Among them, C j is the information content of the jth indicator, S j is the standard deviation of the j-th indicator, R j is the conflict of the jth indicator.

[0056] Calculate the objective weight of the jth indicator as follows:

[0057]

[0058] Among them, W j is the weight of the jth indicator, C jis the information content of the j-th indicator;

[0059] According to the above calculation, the expression of the CRITIC comprehensive evaluation model is:

[0060]

[0061] Among them, ω j is the objective weight of the jth indicator, x i ′ j is the jth data of the ith experimenter after standardization, S i is the comprehensive evaluation score of the i-th experimenter;

[0062] The experimental data were input into the CRITIC comprehensive evaluation model to obtain the comprehensive score of each group of experimenters.

[0063] The dividing line for determining whether elderly drivers are suitable for driving is in the range of [0.453938017,0.532013869].

[0064] In the method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests described in the present invention, in step S3-4, the network in the neural network comprehensive evaluation model is divided into three fully connected layers.

[0065] The three-layer neural network is constructed as follows:

[0066] The input dimension of the first fully connected layer is 13 and the output dimension is 64;

[0067] 13 are the 13 indicators in the evaluation system: eight-word test, observing traffic signs on unfamiliar roads, difficulty dealing with heavy traffic, difficulty dealing with rainy days, blind spot inspection, changes in the position of the vehicle in the driving lane, contact with other road users, speeding or failure to obey traffic lights, eye and head movements, intersection scanning, functional reach test, traffic accidents in the past year, and self-assessment of driving ability;

[0068] The output dimension of the second fully connected layer is 32; the output dimension of the last fully connected layer is 1, and it passes through the sigmoid activation function, and finally the classifier is used to determine whether the test subject is suitable for driving.

[0069] Compared with the prior art, the present invention has the following beneficial effects:

[0070] 1. Construct simulated driving scenarios that reflect the high-risk driving of the elderly, and conduct functional tests and simulated driving tests to make up for the shortcomings of the existing three-force test questions for elderly drivers, which are mostly text description questions. The combination of questionnaire surveys, functional tests and simulated driving tests can fully realize the multi-faceted ability test of elderly drivers' memory, judgment and reaction.

[0071] 2. An evaluation index set is obtained based on the simulated driving experiment data, and a preliminary index set is obtained through expert scoring. The remaining indicators with higher correlation coefficients with low-level results are screened through correlation analysis. As secondary index expansion, it can effectively cover high-risk driving indicators and comprehensively measure the driving suitability of the elderly.

[0072] 3. The experimental data were input into four models, namely, factor analysis, entropy weight-TOPSIS method, CRITIC method, and neural network comprehensive evaluation, and compared with the expert scoring results. The results showed that the model evaluation results were reasonable and effective, and had practical application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 It is a schematic flow chart of the main steps of the method of the present invention;

[0074] Figure 2 This is a schematic diagram of the elderly driver fitness evaluation index system that integrates the three-force test;

[0075] Figure 3 This is an example of a memory test question bank based on traffic signs;

[0076] Figure 4 Sample question bank for a judgement test in bad weather;

[0077] Figure 5 This is an example of a question bank for testing your ability to react to unexpected events. DETAILED DESCRIPTION

[0078] In order to make the purpose and technical solution of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0079] like Figure 1 As shown: A method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests, including:

[0080] (I) Obtain questionnaire survey data on elderly drivers aged 60 and above, conduct descriptive statistical analysis and correlation analysis on the obtained data, and screen out high-risk driving scenarios, including:

[0081] Step S1-1: Design a questionnaire on driving willingness and driving risk of the elderly based on the SP and RP survey methods, distribute the questionnaire online and offline at the same time and collect valid samples;

[0082] Step S1-2: Perform descriptive statistical analysis on the valid sample data to obtain driving scenarios that most elderly drivers choose to avoid and driving scenarios with risk self-assessment scores higher than the average as a high-risk driving scenario set 1;

[0083] Step S1-3: Extract the avoidance choices of elderly drivers for different risk scenarios from the questionnaire, assign weighted scores to the questions to obtain the corresponding total driving risk scores, perform Pearson correlation analysis on the two, and select scenarios with correlation coefficients greater than 0.2 as high-risk driving scenario set 2.

[0084] (II) Construct simulated driving scenarios based on the results of questionnaire survey data analysis, find suitable individuals for simulated driving experiments and functional tests, and obtain experimental data, including:

[0085] Step S2-1: construct four typical driving scenarios and set multiple special events in each driving scenario;

[0086] Among them: four driving scenarios include: urban roads, highways, intersections, and rural roads; special events include: noise interference, heavy traffic flow, driving in rainy days, mixed traffic of motor vehicles and non-motor vehicles, pedestrians rushing out without warning, turning at intersections without signal control, stop and yield signs ahead, and pedestrian crossings ahead;

[0087] Step S2-2: Find a suitable individual to conduct a simulated driving experiment and obtain simulated driving experiment data;

[0088] Step S2-3: guiding the person under test to perform a functional test and obtain functional test data;

[0089] Among them: functional tests include vision tests, physical fitness tests, and cognitive ability tests.

[0090] (III) Based on the simulated driving experiment data, an evaluation index set is obtained, and a driving suitability evaluation system for the elderly is constructed from the perspectives of memory, judgment, and reaction. The system includes:

[0091] Step S3-1: Submit the simulated driving test data of the test subjects to experts for scoring, and select the indicators with obvious differences between the unfit driving group and the fit driving group as the preliminary indicator set;

[0092] Step S3-2: Through correlation analysis, the remaining indicators with high correlation coefficients with the low-level results (such as greater than 0.5) are screened and used as secondary indicator expansion to determine the evaluation indicator set;

[0093] Step S3-3: Divide the evaluation indicators into four categories: memory, judgment, reaction, and comprehensive, and construct an evaluation system for driving suitability of the elderly;

[0094] Among them: Memory: eight-word test;

[0095] Judgment: observing traffic signs on unfamiliar roads, difficulty dealing with heavy traffic, difficulty dealing with rainy days, blind spot inspections, changes in the position of the vehicle in the lane of travel, contact with other road users, speeding or failure to obey traffic lights;

[0096] Responsiveness: Eye and head movements, cross-scanning;

[0097] Comprehensive: functional stretch test, traffic accidents in the past year, self-assessment of driving ability;

[0098] Step S3-4: Input the experimental data into the factor analysis comprehensive evaluation model and compare it with the expert scoring results to verify the rationality of the evaluation system, including:

[0099] Step S3-4-1: Standardize the original data, obtain the common factors, and find the correlation matrix and its eigenvalues ​​and eigenvectors of the standardized data, and calculate the variance contribution rate;

[0100] Step S3-4-2: Calculate the factor loading coefficient after rotation, and linearly combine the original indicators, use regression estimation method and Bartlett estimation method to calculate the scores of each factor, and obtain the linear combination coefficient and weight of each indicator;

[0101] Step S3-4-3: Using the variance contribution rate of each factor as the weight, the comprehensive evaluation index function of the factor analysis method is obtained by the linear combination of each factor:

[0102]

[0103] Among them, ω i is the variance contribution rate of the ith factor, F 1j 、F 2j 、F 3j 、F 4j are the four factors whose characteristic roots of the j-th group of experimental data are greater than 1, F j is the comprehensive evaluation score of the jth experimenter.

[0104] Step S3-4-4: Input the experimental data into the factor analysis comprehensive evaluation function, calculate the comprehensive evaluation score and ranking of each experimenter, and the three with the lowest comprehensive scores are consistent with the three with the lowest ranking in the expert scoring results, which can verify the rationality and effectiveness of the evaluation system. It can be preliminarily considered that the dividing line for determining whether elderly drivers are suitable for driving is within the range of [-1.0208,-0.5576].

[0105] Step S3-5: Input the experimental data into the entropy weight-TOPSIS comprehensive evaluation model and compare it with the expert scoring results to verify the rationality of the evaluation system, including:

[0106] Step S3-5-1: Create an initial matrix:

[0107]

[0108] Among them, (X ij ) m×n are the n evaluation index values ​​of m elderly driver experimenters, i is the experimenter data label, and j is the evaluation index label;

[0109] Step S3-5-2: Perform min-max normalization on the data in the evaluation matrix and convert negative indicators into positive indicators to obtain a normalized matrix:

[0110]

[0111] Among them, (Y ij ) m×n is the normalized value of n evaluation indicators of m experimenters;

[0112] Step S3-5-3: Calculate the information entropy and weight of each indicator, and weight the experimental data to obtain a weighted normalized matrix:

[0113]

[0114] Among them, (V ij ) m×n is the weighted and normalized value of n evaluation indicators of m experimenters;

[0115] Step S3-5-4: Calculate the Euclidean distance from the evaluation object to the positive and negative ideal solutions, and calculate the relative progress:

[0116]

[0117] in, is the distance between the ith solution and the optimal solution, is the distance between the ith solution and the worst solution, C i is the relative progress of the i-th experimenter;

[0118] Step S3-5-5: Input the experimental data into the entropy weight-TOPSIS comprehensive evaluation model to calculate the relative progress and ranking of each experimenter. The three lowest rankings are consistent with the three lowest rankings in the expert scoring results, which can verify the rationality and effectiveness of the evaluation system. It can be preliminarily considered that the dividing line for determining whether elderly drivers are suitable for driving is in the range of [0.514481, 0.520705].

[0119] Step S3-6: Input the experimental data into the CRITIC comprehensive evaluation model and compare it with the expert scoring results to verify the rationality of the evaluation system, including:

[0120] Step S3-6-1: Perform dimensionless processing on the experimental data, calculate the standard deviation and index conflict of each evaluation index, and obtain the information amount calculation formula of each index:

[0121] C j =S j ×R j

[0122] Among them, C j is the information content of the jth indicator, S j is the standard deviation of the j-th indicator, R j is the conflict of the jth indicator.

[0123] Step S3-6-2: Calculate the objective weight of the jth indicator:

[0124]

[0125] Among them, W j is the weight of the jth indicator, C j is the information content of the j-th indicator.

[0126] The expression of the comprehensive evaluation model of CRITIC method is obtained as follows:

[0127]

[0128] Among them, ω j is the objective weight of the jth indicator, x i ′ j is the jth data of the ith experimenter after standardization, S i is the comprehensive evaluation score of the i-th experimenter;

[0129] Step S3-6-3: Input the experimental data into the CRITIC comprehensive evaluation model to obtain the comprehensive score of each group of subjects. The three lowest rankings are consistent with the three lowest rankings in the expert scoring results, which can verify the rationality and effectiveness of the evaluation system. It can be preliminarily considered that the dividing line for determining whether elderly drivers are suitable for driving is in the range of [0.453938017, 0.532013869].

[0130] Step S3-7: Input the experimental data into the neural network comprehensive evaluation model and compare it with the expert scoring results to verify the rationality of the evaluation system, including:

[0131] Step S3-7-1: Build a three-layer neural network, where the input dimension of the first fully connected layer is 13, representing the 13 important indicators in the evaluation system, and the output dimension is 64; the output dimension of the second fully connected layer is 32; the output dimension of the last fully connected layer is 1, and after passing the sigmoid activation function, a classifier is finally used to determine whether the test subject is suitable for driving.

[0132] Step S3-7-2: The number of model training rounds is set to 100 epochs, and the data set is divided into training set and test set using a ratio strategy of 6:4 and 7:3 respectively.

[0133] Step S3-7-3: Input the experimental data into the neural network comprehensive evaluation model to obtain a binary evaluation result. The three subjects identified as unsuitable for driving by the neural network model are consistent with the three lowest rankings in the expert scoring results, which can verify the rationality and effectiveness of the evaluation system.

[0134] (IV) When applied, the driving fitness assessment method for elderly drivers of the present application may include a perception layer, a network layer, a data layer, an application layer and a visualization layer. The data layer is a database, and its functions include data collection and archiving, heterogeneous data integration, abnormal data marking, data storage and data security management.

[0135] The perception layer includes the appointment platform and the cross-departmental cooperation platform. The appointment platform is used to input identity information and the time and place of appointment testing. The cross-departmental cooperation platform can obtain existing driver accident data, medical health data, and social demographic data from other departments through interaction.

[0136] The network layer includes communication equipment, network protocols, security facilities and advanced communication technologies to achieve interconnection and data transmission between devices, and ensure the timeliness and security of transmission.

[0137] The application layer is a driving suitability assessment system for elderly drivers, whose functions include data analysis, indicator extraction, and driving suitability detection.

[0138] The visualization layer includes an elderly-friendly visualization platform and an information query platform. The elderly-friendly visualization platform reduces the cognitive load of the elderly by using large fonts and other designs, and is equipped with elderly-friendly designs such as intelligent voice broadcasting, and provides real-time feedback on test results and driving abnormalities after the test. Information query platforms such as APP, AI assistants, mini-programs, and public accounts synchronize test feedback results in a timely manner and provide query functions.

[0139] Specifically, the driving suitability test proposed in this example includes a test of observing traffic signs on unfamiliar roads. The test questions are as follows: Figure 2 As shown, the person being tested needs to select the traffic signs that appear in the video.

[0140] Specifically, the driving suitability test proposed in this example includes a test for difficult to cope with adverse scenarios, such as Figure 3 As shown, the person being tested needs to select whether the response measures in the severe scenarios in the video are correct or not.

[0141] Specifically, the driving suitability test proposed in this example includes observing the vehicle driving environment test, and the test questions are as follows: Figure 4 As shown, the person being tested needs to select the correct operation in the driving environment in the video.

[0142] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests, characterized in that: include: Step S1: Obtain survey data of elderly drivers over 60 years old through a questionnaire, perform descriptive statistical analysis and Pearson correlation analysis on the obtained data, and screen out high-risk driving scenarios; Step S2: constructing a simulated driving scenario based on the questionnaire survey data analysis results, finding suitable individuals to conduct simulated driving experiments and functional tests, and obtaining simulated driving experiment data; Step S3: Based on the simulated driving experiment data, an evaluation index set is obtained, and a driving suitability evaluation system for the elderly is constructed from the perspectives of memory, judgment, and reaction.

2. The method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests according to claim 1 is characterized in that: The questionnaire survey data of elderly drivers aged 60 and above is obtained, and descriptive statistical analysis and Pearson correlation analysis are performed on the obtained data to screen out high-risk driving scenarios, including: Step S1-1: Design a questionnaire on driving willingness and driving risk of the elderly based on the SP and RP survey methods, distribute the questionnaire online and offline at the same time and collect valid samples; Step S1-2: Perform descriptive statistical analysis on the valid sample data to obtain driving scenarios that most elderly drivers choose to avoid and driving scenarios with risk self-assessment scores higher than the average as a high-risk driving scenario set 1; Step S1-3: extract the avoidance choices of elderly drivers for different risk scenarios from the questionnaire, assign weighted scores to the questions and obtain the corresponding total driving risk scores, perform Pearson correlation analysis on the two, and select scenarios with correlation coefficients higher than 0.2 as high-risk driving scenario set 2; Step S1-4: Based on the high-risk driving scenario set 1 obtained in step S1-2, the high-risk driving scenario set 2 obtained in step S1-3 is used as a subsequent simulation test scenario.

3. The method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests according to claim 1 is characterized in that: The simulated driving scenario is constructed based on the questionnaire survey data analysis results, suitable individuals are found for simulated driving experiments and functional tests, and experimental data are obtained, including: Step S2-1: construct four typical driving scenarios, and set multiple special events in each driving scenario; the four driving scenarios include: urban road section, highway, intersection, and rural road section; Special events include: noise interference, heavy traffic flow, driving in rainy weather, mixed traffic of motor vehicles and non-motor vehicles, pedestrians rushing out without warning, turning at intersections without signal control, and stop and yield signs ahead; Step S2-2: Find a suitable individual to conduct a simulated driving experiment and obtain simulated driving experiment data; Step S2-3: guiding the person under test to perform a functional test and obtain functional test data; Functional tests include: vision test, physical fitness test, and cognitive ability test.

4. A method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests according to claim 1 or 3, characterized in that: The evaluation index set is obtained based on the simulated driving experiment data and divided into three categories: memory, judgment, and reaction, so as to construct an evaluation system for driving suitability of the elderly, including: Based on high-risk driving scenario set 1 and high-risk driving scenario set 2, simulated driving scenarios were built on the computer using software and built into the laboratory's simulated driving device for the tester's on-site simulated driving experiment; Step S3-1: The simulated driving test data of the test subjects are given to experts for scoring and screening, and the indicators with obvious differences between the unfit driving group and the fit driving group are screened out as the preliminary indicator set; Step S3-2: Screen out indicators with high correlation with the results of unsuitable driving samples from the remaining data information covered by the questionnaire, experiment and functional test through correlation analysis, and use them as secondary indicator expansion to determine the evaluation indicator set; Step S3-3: taking the preliminary index set in step S3-1 and the evaluation index set expanded in step S3-2 as the final evaluation index; dividing the final evaluation index into four categories: memory, judgment, reaction, and comprehensive, and constructing an evaluation system for driving suitability of the elderly; Step S3-4: According to the obtained evaluation system, the data obtained from the questionnaire is applied to four models: factor analysis, entropy weight-TOPSIS method, CRITIC method, and neural network comprehensive evaluation, to verify the rationality of the evaluation system from the perspective of binary classification and score ranking.

5. The method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests according to claim 4 is characterized in that: In the step S3-3, Evaluation indicators focus on memory including: eight-word test; Judgment includes: observing traffic signs on unfamiliar roads, difficulty dealing with heavy traffic, difficulty dealing with rainy days, blind spot inspections, changes in the vehicle's position in the lane of travel, contact with other road users, speeding or failure to obey traffic lights; Responsiveness includes: eye and head movements, cross-scanning; Comprehensive includes: functional stretch test, traffic accidents in the past year, and self-assessment of driving ability.

6. The method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests according to claim 4 is characterized in that: In step S3-4, the comprehensive evaluation model expression of the factor analysis method is: Among them, ω i is the variance contribution rate of the i-th factor, F 1j 、F 2j 、F 3j 、F 4j are the four factors whose characteristic roots of the j-th group of experimental data are greater than 1, F j is the comprehensive evaluation score of the jth experimenter.

7. The method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests according to claim 4 is characterized in that: In step S3-4, the steps of establishing the entropy weight-TOPSIS method comprehensive evaluation model are as follows: Create the initial matrix: Among them, (X ij ) m×n are the n evaluation index values ​​of m elderly driver experimenters, i is the experimenter data label, and j is the evaluation index label; Perform min-max normalization on the data in the evaluation matrix and convert negative indicators into positive indicators to obtain the normalized matrix: Among them, (Y ij ) m×n is the normalized value of n evaluation indicators of m experimenters; Calculate the information entropy and weight of each indicator, and weight the experimental data to obtain the weighted normalized matrix: Among them, (V ij ) m×n is the weighted and normalized value of n evaluation indicators of m experimenters; Calculate the Euclidean distance from the evaluation object to the positive and negative ideal solutions, and calculate the relative progress: in, is the distance between the ith solution and the optimal solution, is the distance between the ith solution and the worst solution, C i is the relative progress of the i-th experimenter; The experimental data were input into the entropy weight-TOPSIS comprehensive evaluation model to calculate the score of each experimenter. The dividing line for determining whether the elderly driver is suitable for driving is in the range of [0.514481, 0.520705].

8. The method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests according to claim 4 is characterized in that: In step S3-4, the steps for establishing the CRITIC comprehensive evaluation model are as follows: The experimental data was dimensionless processed, the standard deviation and conflict of each evaluation index were calculated, and the information amount calculation formula of each index was obtained: C j =S j ×R j Among them, C j is the information content of the jth indicator, S j is the standard deviation of the j-th indicator, R j is the conflict of the jth indicator. Calculate the objective weight of the jth indicator as follows: Among them, W j is the weight of the jth indicator, C j is the information content of the j-th indicator; According to the above calculation, the expression of the CRITIC comprehensive evaluation model is: Among them, ω j is the objective weight of the jth indicator, x i ′ j is the jth data of the ith experimenter after standardization, S i is the comprehensive evaluation score of the i-th experimenter; The experimental data were input into the CRITIC comprehensive evaluation model to obtain the comprehensive score of each group of experimenters. The dividing line for determining whether elderly drivers are suitable for driving is in the range of [0.453938017,0.532013869].

9. The method for constructing a driving fitness evaluation system for elderly drivers integrating three-force tests according to claim 4 is characterized in that: In step S3-4, the network in the neural network comprehensive evaluation model is divided into three fully connected layers. The three-layer neural network is constructed as follows: The input dimension of the first fully connected layer is 13 and the output dimension is 64; 13 are the 13 indicators in the evaluation system: eight-word test, observing traffic signs on unfamiliar roads, difficulty dealing with heavy traffic, difficulty dealing with rainy days, blind spot inspection, changes in the position of the vehicle in the driving lane, contact with other road users, speeding or failure to obey traffic lights, eye and head movements, intersection scanning, functional reach test, traffic accidents in the past year, and self-assessment of driving ability; The output dimension of the second fully connected layer is 32; the output dimension of the last fully connected layer is 1, and it passes through the sigmoid activation function, and finally the classifier is used to determine whether the test subject is suitable for driving.

Images