Automatic driving passenger car in-transit service configuration method based on user portrait

Abstract

The invention discloses an automatic driving passenger car in-transit service configuration method based on a user portrait, and relates to the technical field of automatic driving. According to the method, an initial influence factor set is constructed through literature retrieval and questionnaire survey, and significant influence factors are screened in combination with a multiple linear regression model; integrating the significant factors and driving factor data collected by a driving simulator, and determining the level classification number and threshold standard of each factor by using a clustering algorithm; based on basic information, travel habits and preference data of users, passenger user portraits are described through potential profile analysis, and scientific classification of user groups is realized; and finally, according to a classification result and a threshold standard, personalized in-transit service schemes are automatically configured for different users. According to the method, through quantitative screening of influence factors, dynamic clustering and threshold division and accurate user portraits, the problems of high subjectivity and low adaptability of a service scheme in a traditional method are solved, and the matching efficiency of services and demands and the passenger satisfaction degree are remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of autonomous driving, and more particularly, to a method for configuring in-vehicle services for autonomous driving passenger cars based on user portraits. Background Art

[0002] With the rapid development of autonomous driving technology, the in-vehicle service requirements of passenger cars are increasingly showing a trend of diversification and personalization. During the use of autonomous driving passenger cars, users' requirements for in-vehicle services are affected by various factors, including but not limited to environmental factors, driving factors, service factors, and travel plan factors. The diversity of these factors leads to differences in user requirements.

[0003] Traditional service solutions often adopt a unified and fixed mode. Although this mode is simple in design and implementation, it cannot effectively meet the personalized needs of different user groups, resulting in many deficiencies in the user's riding experience. In addition, with the development of big data and artificial intelligence technologies, user portrait technology has gradually matured. By collecting and analyzing data such as users' personal information, travel habits, and preference settings, a detailed and accurate user portrait can be constructed, providing strong data support for the configuration of personalized service solutions.

[0004] However, there is currently a lack of an effective method that can make full use of user portrait technology to configure personalized in-vehicle service solutions for autonomous driving passenger cars. Therefore, developing a method for configuring in-vehicle services for autonomous driving passenger cars based on user portraits has become an urgent need for the industry's development. Summary of the Invention

[0005] (I) Technical Problems to be Solved

[0006] In view of the problems in the related art, the present invention provides a method for configuring in-vehicle services for autonomous driving passenger cars based on user portraits. By constructing user portraits, analyzing user characteristics and needs, and configuring personalized in-vehicle service solutions for autonomous driving passenger cars, the riding experience and satisfaction of users can be improved.

[0007] (II) Technical Solutions

[0008] To solve the above technical problems, the present invention is implemented through the following technical solutions:

[0009] The present invention discloses a method for configuring in-vehicle services for autonomous driving passenger cars based on user portraits, including the following steps:

[0010] S1. Construct an initial set of influencing factors; design a questionnaire based on the initial set of influencing factors; conduct a questionnaire survey using the questionnaire to obtain valid questionnaires;

[0011] S2. Construct a multiple linear regression model; combine the valid questionnaires with the multiple linear regression model to screen each influencing factor in the initial set of influencing factors, and obtain a set of significant influencing factors;

[0012] S3. Collect the preference results of the set of significant influencing factors to obtain the final preference result data; according to the final preference result data and combined with the clustering algorithm, obtain the classification threshold standard set for the influencing factor levels of the in-vehicle service plan for autonomous driving passenger cars;

[0013] S4. According to the classification threshold standard set, depict the user portraits of passengers of autonomous driving passenger cars in the valid questionnaires; based on the user portraits of passengers of autonomous driving passenger cars, configure the in-vehicle service plan.

[0014] Preferably, the S1 includes the following steps:

[0015] S11. Through literature retrieval, comprehensively analyze the influencing factors confirmed in previous studies on the in-vehicle service plan for passenger cars, preliminarily screen the influencing factors for the in-vehicle service plan for autonomous driving passenger cars, and construct an initial set of influencing factors a = {a1, a2,..., a i ,...a n}, where a i represents the i-th influencing factor of the in-vehicle service plan for autonomous driving passenger cars, and n represents the total number of influencing factors; the influencing factors include environmental factors such as in-vehicle temperature, humidity, odor, and noise; driving factors such as triaxial acceleration change, speed, and driving smoothness; service factors such as seating space size and entertainment system; travel plan factors such as travel time and travel cost;

[0016] S12. Design a questionnaire for passengers of autonomous driving passenger cars for the influencing factors in the initial set of influencing factors; the questionnaire includes basic information (age, gender, occupation, income situation), travel habits (travel frequency, common time period, destination type), users' preferences for each influencing factor (such as what is your favorite in-vehicle temperature), the comprehensive satisfaction score of the virtual in-vehicle service plan formed by the combination of each influencing factor (such as in the scenario of temperature 26 degrees, humidity 40%, etc., how satisfied are you with the above in-vehicle service plan? 1 is very dissatisfied, 10 is very satisfied), and the preference changes of each influencing factor under different scenarios (peak / off-peak hours, different seasons, weekend / weekday, multiple people traveling (with the elderly, children)) (such as how does your preference for in-vehicle temperature change during peak hours, 1 is a significant decrease, 5 is a significant increase);

[0017] S13. Adopt the methods of online questionnaire survey and offline questionnaire survey of passengers of autonomous driving passenger cars to collect a certain number of questionnaire results, eliminate the invalid questionnaires with missing data and the same answer options filled in, and obtain the valid questionnaires;

[0018] Through the above steps of literature retrieval and multi-dimensional questionnaire design, the initial set of influencing factors for the in-transit service plan of autonomous passenger vehicles was systematically constructed, and based on the questionnaire survey combining online and offline, user preferences and scenario-based demand data were efficiently obtained; ensuring the scientificity and comprehensiveness of the influencing factors and avoiding subjective omissions; accurately capturing the dynamic demands of users for the service plan through structured questionnaires, providing data support for personalized services; eliminating invalid questionnaires to ensure the reliability of the analysis results and laying a high-quality data foundation for subsequent modeling and user portrait construction.

[0019] Preferably, the S2 includes the following steps:

[0020] S21. Construct a multiple linear regression model;

[0021] S22. Combine the multiple linear regression model to screen each influencing factor in the initial set of influencing factors to obtain a set of significant influencing factors;

[0022] Through the above steps of constructing a multiple linear regression model, the initial influencing factors are associated with the user satisfaction scores, quantitatively analyzing the influence degree of each factor on the service plan, providing accurate input for subsequent level classification and user portrait, and ultimately enhancing the scientificity and operability of the service plan.

[0023] Preferably, the S21 includes the following steps:

[0024] S211. Set the significance level as α, take each influencing factor in the initial set of influencing factors as the independent variable, and take the satisfaction scores of the passengers in the valid questionnaires as the dependent variable to construct a multiple linear regression model, perform parameter estimation, and obtain the regression coefficients of each independent variable and the values of the error terms. The formula is as follows,

[0025] Y = β0 + β1X1 + β2X2 +... + β n X n + ε;

[0026] where Y represents the satisfaction score; β0 represents the intercept term; β1, β2,... β n are the regression coefficients of a1, a2... a n indicating the degree of influence of each independent variable on the dependent variable; ε represents the error term;

[0027] Through the above steps of setting the significance level, taking the initial influencing factors as the independent variables and the passenger satisfaction scores as the dependent variables, constructing a multiple linear regression model and performing parameter estimation, the regression coefficients of each independent variable and the values of the error terms are obtained; it can effectively identify and quantify the key factors affecting passenger satisfaction.

[0028] Preferably, S22 includes the following steps:

[0029] S221. For each independent variable, perform a t-test using SPSS software to obtain a set of P-values for the independent variables P = {P1, P2,..., P i ,...P n}, where P i represents the P-value of the i-th independent variable;

[0030] S222. For the independent variables with P-values lower than the significance level α in the set of P-values of independent variables, remove these independent variables from the multiple linear regression model, refit the multiple linear regression model, calculate the new regression coefficients and error terms; calculate the P-values of each independent variable again;

[0031] S223. Repeat S221 and S222. When the P-values of all independent variables are less than the significance level α, stop the iteration to obtain a set of significant influencing factors b = {b1, b2,..., b i ,...b m}, where b i represents the i-th influencing factor with a P-value less than the significance level α, and m represents the total number of influencing factors with P-values less than the significance level α;

[0032] Otherwise, continue to remove independent variables, perform fitting of the multiple linear regression model, calculate the regression coefficients, error terms and P-values of independent variables until the P-values of all independent variables are less than the significance level α;

[0033] The above steps dynamically remove non-significant factors through and based on significance tests, and iteratively optimize the model; screen key influencing factors through data-driven methods to avoid interference from redundant variables and improve the interpretability and prediction accuracy of the model; significance tests ensure the objectivity and reliability of the screening results, providing a statistical basis for service plan configuration; the determination of the set of significant influencing factors focuses on the real needs of users, reduces the consumption of computing resources for invalid parameter analysis, provides accurate input for subsequent level classification and user profiling, and ultimately enhances the scientificity and operability of the service plan.

[0034] Preferably, S3 includes the following steps:

[0035] S31. Collect the preference results of the influencing factors obtained through valid questionnaires in the set of significant influencing factors, as well as the preference results of the driving factors not obtained through valid questionnaires, to obtain the final preference results;

[0036] S32. Use the final preference result data and combine it with the clustering algorithm to obtain a set of classification threshold criteria for the influencing factor levels of the in-vehicle service plan for autonomous passenger vehicles;

[0037] The above steps obtain the final preference result by collecting the preference results of the significant influence set; formulate a set of hierarchical classification threshold criteria for the influencing factors of the in-vehicle service plan of autonomous passenger vehicles; refine the key factors affecting the in-vehicle service of autonomous passenger vehicles and also provide clear criteria for the classification of their importance levels; this helps service providers better understand and meet passenger needs, optimize service plans, and improve passenger satisfaction and riding experience.

[0038] Preferably, S31 includes the following steps:

[0039] S311. Sort out the preference result data of the influencing factors in the significant influence factor set by users in the valid questionnaires to obtain questionnaire preference result data; collect the satisfaction score data of users under the change of each driving factor through a driving simulator to obtain driving factor preference result data; the driving factors include triaxial acceleration change, speed, and driving smoothness.

[0040] S312. Integrate the questionnaire preference result and the driving factor preference result to obtain the final preference result data.

[0041] The above steps integrate the preference data in the user questionnaire and the driving factor score data collected by the driving simulator, and fuse them to form the final preference result covering the subjective and objective demand characteristics; combine the user's subjective preference and the objective data of the driving scenario to improve the comprehensiveness and accuracy of the influencing factor analysis; through multi-source data complementarity (questionnaire static preference + simulator dynamic scoring), enhance the adaptability and scenario coverage ability of the service plan configuration; provide a high-confidence data basis for subsequent clustering analysis to ensure that the hierarchical classification threshold can truly reflect the combined influence of user needs and driving scenarios, thus supporting the design of highly personalized service plans.

[0042] Preferably, S32 includes the following steps:

[0043] S321. Use a clustering algorithm to perform clustering analysis on the final preference result data to obtain the optimal clustering number set d = {d1, d2,..., d i ,...d m}; where d i represents the optimal clustering number of the i-th influencing factor of the in-vehicle service plan of autonomous passenger vehicles; the optimal clustering number represents the number of riding preferences corresponding to each influencing factor in the significant influence factor set; for example, if the optimal clustering number of the in-vehicle temperature preference data is 5, then the in-vehicle temperature is divided into 5 levels, namely, level 1, level 2, level 3, level 4, and level 5.

[0044] S322. Each cluster in the optimal number of clusters represents a set of samples with similar ride preference characteristics. Calculate the distance from the samples in each cluster to the center point of its belonging cluster to obtain a set of classification threshold criteria \(e = \{e_1, e_2, \ldots, e i , \ldots, e m \}\); where \(e i represents the thresholds corresponding to various ride preferences of each influencing factor; for example, the thresholds corresponding to 5 levels of in-vehicle temperature are level 1 (21 - 23.5 degrees), level 2 (23.5 - 25 degrees), level 3 (25 degrees), level 4 (25 - 27 degrees), and level 5 (27 - 28 degrees).

[0045] The above steps analyze the preference characteristics of the final preference result data through a clustering algorithm, determine the optimal number of classifications for each factor, and set the level thresholds based on the distance distribution between the samples and the cluster centers; quantify the abstract user preferences into specific and operable classification criteria, solving the problem of fuzzy service parameters; dynamically divide the thresholds through data-driven to avoid subjective preset biases and adapt to the diverse needs of different user groups; provide standardized input for subsequent user profiling and personalized service configuration, ensuring that the service plan accurately matches the real preferences of users in different scenarios, significantly improving the service response efficiency and passenger experience of autonomous driving passenger vehicles.

[0046] Preferably, the said S4 includes the following steps:

[0047] S41. According to the set of classification threshold criteria, depict the user profiles of passengers of autonomous driving passenger vehicles in valid questionnaires to obtain the optimal number of classifications;

[0048] S42. Analyze the user group characteristics of each category and the corresponding in-transit service plan results under the optimal number of classifications, and store the user group characteristics and in-transit service plans of each category; after the users of this category place an order for an autonomous driving passenger vehicle, configure the in-transit service plan according to historical data.

[0049] The above steps depict the user profiles of passengers of autonomous driving passenger vehicles in valid questionnaires by using the set of classification threshold criteria to determine the optimal number of classifications; analyze the characteristics of each category of user groups and their corresponding in-transit service plans, and store this information; when users of the corresponding category place an order, the system can configure a suitable in-transit service plan for them according to historical data; realizing the precise classification and personalized service configuration of passengers of autonomous driving passenger vehicles.

[0050] Preferably, the said S41 includes the following steps:

[0051] S411. Sort out the basic information and travel habit data in the aforementioned valid questionnaires, and combine with the classification threshold standard set to convert the questionnaire data into structured feature vector data. For example, User 1: {Age: 30; Gender: Male; Occupation: Senior executive of a private enterprise; Income situation: High; Travel frequency: High; Common time period: Peak period; Destination type: Office building; Preferred in-vehicle temperature: Low; Preferred in-vehicle humidity: High};

[0052] S412. Use latent profile analysis in Mplus software to classify the users in the structured feature vector data;

[0053] Set the number of classifications k = 1, and gradually increase the value of k. By observing the AIC, BIC, ABIC, entropy value, likelihood ratio test index LMR, and Bootstrap-based likelihood ratio test index BLRT, obtain the optimal number of classifications. Among them, AIC, BIC, and ABIC are information evaluation indicators, and the smaller the values of the three indicators, the better the model fitting; the entropy value represents the classification accuracy, and the closer it is to 1, the more accurate the classification. When it is less than 0.6, it indicates that more than 20% of the individuals have classification errors, and when it is greater than 0.8, it indicates that the classification accuracy exceeds 90%; the p-value of the LMR and BLRT test values reaching the significant level (P < 0.05) indicates that the model with k categories is better than the model with k - 1 categories;

[0054] The above steps realize the precise division of the passenger group by converting the user multi-dimensional data into structured feature vectors and dynamically optimizing the user classification model based on latent profile analysis; by standardizing data conversion (such as mapping preferences to classification thresholds), eliminate the data dimension differences, and improve the compatibility and efficiency of portrait analysis; combine statistical indicators such as AIC, BIC, and entropy value to scientifically determine the optimal number of classifications, avoid the subjective deviation of artificial preset classification, identify the core characteristics of different user groups based on high-precision classification (such as "high-income business users prefer quiet environments"), provide a quantifiable configuration basis for personalized service solutions, significantly enhance the adaptability and user satisfaction of autonomous driving passenger car services, and at the same time reduce the service trial-and-error cost.

[0055] (III) Beneficial effects

[0056] The present invention has the following beneficial effects:

[0057] The present invention more comprehensively, scientifically, and accurately constructs the influence factor set and its hierarchical classification threshold standard of the in-vehicle service plan of autonomous driving passenger cars through literature retrieval pre-screening, quantitative influence factor significance test, and hierarchical classification.

[0058] Through massive multi-dimensional data collection and latent profile analysis, scientific classification of autonomous driving passenger vehicle users is carried out, and a comprehensive and accurate user portrait is constructed by analyzing the characteristics and riding preference data of various types of users, providing theoretical support for the provision of personalized in-vehicle service solutions for autonomous driving passenger vehicles; realizing the precise matching of services and demands and improving the efficiency of service provision.

[0059] Through the configuration method based on the user portrait, the present invention can provide highly personalized in-vehicle service solutions for the passengers of autonomous driving passenger vehicles, significantly enhancing the riding experience and satisfaction of passengers and meeting the needs of modern consumers for personalized services.

[0060] Of course, it is not necessary for any product implementing the present invention to achieve all the above-mentioned advantages simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings without creative efforts.

[0062] Figure 1 It is a schematic flow chart of a method for configuring in-vehicle services of an autonomous driving passenger vehicle based on a user portrait according to the present invention;

[0063] Figure 2 It is a schematic flow chart of determining a set of significant influencing factors in a method for configuring in-vehicle services of an autonomous driving passenger vehicle based on a user portrait according to the present invention;

[0064] Figure 3 It is a schematic flow chart of obtaining user portrait classification in a method for configuring in-vehicle services of an autonomous driving passenger vehicle based on a user portrait according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0065] The technical solutions in the embodiments of the invention will be clearly and completely described below with reference to the drawings in the embodiments of the invention. Obviously, the described embodiments are only a part of the embodiments of the invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the invention belong to the scope of protection of the invention.

[0066] In the description of the present invention, it should be understood that the terms "opening", "upper", "lower", "top", "middle", "inner", etc. indicating the orientation or position relationship are only for the convenience of describing the invention and simplifying the description, rather than indicating or implying that the components or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation to the invention.

[0067] Example 1:

[0068] Please refer to Figure 1 、 Figure 2 、 Figure 3 , the present invention discloses a method for configuring in - transit services for an autonomous driving passenger vehicle based on a user profile, including the following steps:

[0069] S1. Construct an initial set of influencing factors; design a questionnaire based on the initial set of influencing factors; conduct a questionnaire survey using the questionnaire to obtain valid questionnaires;

[0070] The said S1 includes the following steps:

[0071] S11. Through literature retrieval and comprehensive analysis of the influencing factors confirmed in previous studies on in - transit service plans for passenger vehicles, preliminarily screen the influencing factors for in - transit service plans for autonomous driving passenger vehicles, and construct an initial set of influencing factors \(a=\{a_1,a_2,\cdots,a i ,\cdots,a n \}\), where \(a i \) represents the \(i\) - th influencing factor of the in - transit service plan for an autonomous driving passenger vehicle, and \(n\) represents the total number of influencing factors; the influencing factors include environmental factors such as in - vehicle temperature, humidity, odor, and noise; driving factors such as tri - axis acceleration change, speed, and driving smoothness; service factors such as seating space size and entertainment system; travel plan factors such as travel time and travel cost;

[0072] S12. Design a questionnaire for passengers of autonomous driving passenger vehicles with the influencing factors in the initial set of influencing factors; the questionnaire includes basic information (age, gender, occupation, income situation), travel habits (travel frequency, common time period, destination type), user preferences for each influencing factor (such as what temperature do you like most in the vehicle), comprehensive satisfaction scores for virtual in - transit service plans formed by combinations of each influencing factor (such as in scenarios with a temperature of 26 degrees and a humidity of 40%, how satisfied are you with the above in - transit service plan as a whole? 1 means very dissatisfied, 10 means very satisfied), and preference changes for each influencing factor in different scenarios (peak / off - peak hours, different seasons, weekend / weekday, multiple - person travel (with the elderly, children)) (such as how do you prefer the in - vehicle temperature to change during peak hours, 1 means a significant decrease, 5 means a significant increase)

[0073] S13. Adopt online questionnaire surveys and offline questionnaire surveys of passengers of autonomous driving passenger vehicles to collect a certain number of questionnaire results, eliminate invalid questionnaires with missing data and those with the same answer options filled in, and obtain valid questionnaires;

[0074] S2. Construct a multiple linear regression model; combine the valid questionnaires with the multiple linear regression model to screen each influencing factor in the initial set of influencing factors to obtain a set of significant influencing factors;

[0075] The S2 includes the following steps:

[0076] S21. Construct a multiple linear regression model;

[0077] The S21 includes the following steps:

[0078] S211. Set the significance level as α. Take each influencing factor in the initial set of influencing factors as an independent variable, and take the satisfaction score of the passengers in the valid questionnaires as the dependent variable to construct a multiple linear regression model, perform parameter estimation to obtain the regression coefficients of each independent variable and the values of the error terms. The formula is as follows.

[0079] Y = β0 + β1X1 + β2X2 +... + β n X n + ε;

[0080] Among them, Y represents the satisfaction score; β0 represents the intercept term; β1, β2,... β n are the regression coefficients of a1, a2... a n , representing the degree of influence of each independent variable on the dependent variable; ε represents the error term;

[0081] S22. Combine the multiple linear regression model to screen each influencing factor in the initial set of influencing factors to obtain a set of significant influencing factors;

[0082] The S22 includes the following steps:

[0083] S221. For each independent variable, use SPSS software to perform a t-test to obtain a set of P-values of the independent variables P = {P1, P2,..., P i ,... P n}, where P i represents the P-value of the i-th independent variable;

[0084] S222. For the independent variables in the set of P-values of the independent variables whose P-values are lower than the significance level α, remove this independent variable from the multiple linear regression model, refit the multiple linear regression model, calculate the new regression coefficients and error terms; calculate the P-values of each independent variable again;

[0085] S223. Repeat S221 and S222. When the P-values of all independent variables are less than the significance level α, stop the iteration to obtain a set of significant influencing factors b = {b1, b2,..., b i ,... b m}, where bi It represents the influencing factor whose i-th P value is less than the significance level α, and m represents the total number of influencing factors whose P values are less than the significance level α;

[0086] Otherwise, continue to eliminate independent variables and perform fitting of the multiple linear regression model, calculate the regression coefficients, error terms, and P values of the independent variables until the P values of all independent variables are less than the significance level α;

[0087] S3. Collect the preference results of the set of significant influencing factors to obtain the final preference result data; according to the final preference result data and combined with the clustering algorithm, obtain the classification threshold standard set of the influencing factor levels of the in-transit service plan for autonomous driving passenger vehicles;

[0088] The S3 includes the following steps:

[0089] S31. In the set of significant influencing factors, collect the preference results of the influencing factors obtained through valid questionnaires and the preference results of the driving factors not obtained through valid questionnaires to obtain the final preference results;

[0090] The S31 includes the following steps:

[0091] S311. Sort out the preference result data of the influencing factors in the set of significant influencing factors by users in the valid questionnaires to obtain the questionnaire preference result data; collect the satisfaction score data of users under the changes of various driving factors through the driving simulator to obtain the driving factor preference result data; the driving factors include triaxial acceleration change, speed, and driving smoothness;

[0092] S312. Integrate the questionnaire preference results and the driving factor preference results to obtain the final preference result data;

[0093] S32. Use the final preference result data and combined with the clustering algorithm to obtain the classification threshold standard set of the influencing factor levels of the in-transit service plan for autonomous driving passenger vehicles;

[0094] The S32 includes the following steps:

[0095] S321. Use the clustering algorithm to perform clustering analysis on the final preference result data to obtain the optimal clustering number set d = {d1, d2,..., d i ,...d m}; where d i represents the optimal clustering number of the i-th influencing factor of the in-transit service plan for autonomous driving passenger vehicles; the optimal clustering number represents the number of riding preferences corresponding to each influencing factor in the set of significant influencing factors; if the optimal clustering number of the in-vehicle temperature preference data is 5, then the in-vehicle temperature is divided into 5 levels, namely, first level, second level, third level, fourth level, and fifth level;

[0096] S322. Each cluster in the optimal number of clusters represents a group of samples with similar ride preference characteristics. Calculate the distance from the samples in each cluster to the center point of its belonging cluster to obtain the classification threshold standard set \(e = \{e_1, e_2,..., e i ,...e m \}; where \(e i \) represents the thresholds corresponding to various ride preferences of each influencing factor; for example, the thresholds corresponding to 5 levels of in-vehicle temperature are level 1 (21 - 23.5 degrees), level 2 (23.5 - 25 degrees), level 3 (25 degrees), level 4 (25 - 27 degrees), and level 5 (27 - 28 degrees);

[0097] S4. Characterize the user portraits of passengers in autonomous driving passenger cars in the valid questionnaires according to the classification threshold standard set; configure the in-transit service plan based on the user portraits of passengers in autonomous driving passenger cars.

[0098] The S4 includes the following steps:

[0099] S41. Characterize the user portraits of passengers in autonomous driving passenger cars in the valid questionnaires according to the classification threshold standard set to obtain the optimal number of classifications;

[0100] The S41 includes the following steps:

[0101] S411. Sort out the basic information and travel habit data in the aforementioned valid questionnaires, and combine the classification threshold standard set to convert the questionnaire data into structured feature vector data; for example, User 1: {Age: 30; Gender: Male; Occupation: Senior executive of a private enterprise; Income situation: High; Travel frequency: High; Common time period: Peak period; Destination type: Office building; Preferred in-vehicle temperature: Lower; Preferred in-vehicle humidity: Higher};

[0102] S412. Use latent profile analysis in Mplus software to classify the data in the structured feature vector data for user portraits;

[0103] Set the number of classifications \(k = 1\) and gradually increase the value of \(k\). By observing the AIC, BIC, ABIC, entropy value, likelihood ratio test index LMR, and bootstrap-based likelihood ratio test index BLRT; obtain the optimal number of classifications;

[0104] S42. Analyze the user group characteristics of each category and the corresponding in-transit service plan results under the optimal number of classifications, and store the user group characteristics and in-transit service plans of each category; after the users in this category place an order for an autonomous driving passenger car, configure the in-transit service plan according to the historical data.

[0105] In the description of this specification, the description referring to terms such as "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0106] The preferred embodiments of the invention disclosed above are only used to help explain the invention. The preferred embodiments do not exhaust all the details and do not limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principle and practical application of the invention, so that those skilled in the art can understand and utilize the invention well.

Claims

1. A method for configuring on-the-way services for autonomous driving passenger vehicles based on user portraits, characterized in that: The following steps are involved: S1. Construct an initial set of influencing factors; design a questionnaire based on the initial set of influencing factors; conduct a questionnaire survey using the questionnaire to obtain a valid questionnaire; S2. Construct a multiple linear regression model; combine the effective questionnaire with the multiple linear regression model to screen each influencing factor in the initial influencing factor set to obtain a significant influencing factor set; S3, collect the preference results of the significant influencing factor set to obtain the final preference result data; Based on the final preference result data and combined with the clustering algorithm, a set of classification threshold standards for the level of influencing factors of the on-the-way service plan for autonomous driving passenger vehicles is obtained; S4. Characterize the user portrait of autonomous driving passenger cars in the valid questionnaire according to the classification threshold standard set; configure the on-the-way service plan based on the user portrait of autonomous driving passenger cars.

2. The method for configuring on-the-way services for autonomous driving passenger vehicles based on user portraits according to claim 1, characterized in that: The S1 comprises the following steps: S11. Through literature retrieval, comprehensively analyze the influencing factors that have been confirmed in previous studies to affect the on-the-way service plan of passenger vehicles, preliminarily screen the influencing factors for the on-the-way service plan of autonomous driving passenger vehicles, and construct the initial influencing factor set a = {a1, a2, ..., a i ,...a n }, where a i represents the i-th influencing factor of the on-the-way service plan of the autonomous driving passenger car, and n represents the total number of influencing factors; S12. Based on the influencing factors in the initial influencing factor set, design a questionnaire for autonomous driving passenger car passengers containing basic information, travel habits, user preferences for each influencing factor, a comprehensive satisfaction score of the virtual in-transit service solution formed by the combination of each influencing factor, and changes in preferences for each influencing factor in different scenarios; S13. Collect a certain number of questionnaire results by conducting online questionnaire surveys and offline questionnaire surveys on passengers of autonomous driving vehicles, eliminate invalid questionnaires with missing data and the same answer options, and obtain valid questionnaires.

3. The method for configuring on-the-way services for autonomous driving passenger vehicles based on user portraits according to claim 1, characterized in that: The S2 comprises the following steps: S21. Construct a multiple linear regression model; S22. Combined with the multiple linear regression model, each influencing factor in the initial influencing factor set is screened to obtain a set of significant influencing factors.

4. The method for configuring on-the-way services for autonomous driving passenger vehicles based on user portraits according to claim 3, characterized in that: The S21 comprises the following steps: S211. Set the significance level to α, take each influencing factor in the initial influencing factor set as an independent variable, take the passenger satisfaction score in the valid questionnaire as a dependent variable, construct a multivariate linear regression model, perform parameter estimation, and obtain the regression coefficient of each variable and the error term value. The formula is as follows: Y = β0 + β1X1 + β2X2 + ... + β n X n +ε; Among them, Y represents the satisfaction score; β0 represents the intercept term; β1, β2, ...β n for a1, a2...a n The regression coefficient indicates the degree of influence of each independent variable on the dependent variable; ε represents the error term.

5. The method for configuring on-the-way services for autonomous driving passenger vehicles based on user portraits according to claim 3, characterized in that: The S22 comprises the following steps: S221, for each independent variable, use SPSS software to perform t test, and obtain the independent variable P value set P = {P1, P2, ..., P i ,...P n }, where P i represents the P value of the ith independent variable; S222. For the independent variables whose P values ​​in the independent variable P value set are lower than the significance level α, the independent variables are removed from the multiple linear regression model, and the multiple linear regression model is refitted to calculate the new regression coefficient and error term; and the P value of each independent variable is calculated again; S223, repeat S221 and S222. When the P values ​​of all independent variables are less than the significance level α, stop the iteration and obtain the significant influencing factor set b = {b1, b2, ..., b i ,...b m }, where b i represents the influencing factor whose P value is less than the significance level α, and m represents the total number of influencing factors whose P value is less than the significance level α; Otherwise, continue to eliminate independent variables and perform multivariate linear regression model fitting, calculate regression coefficients, error terms, and independent variable P values ​​until the P values ​​of all independent variables are less than the significance level α.

6. The method for configuring on-the-way services for autonomous driving passenger vehicles based on user portraits according to claim 1, characterized in that: The S3 comprises the following steps: S31, collecting preference results of the influencing factors obtained through valid questionnaires and preference results of the driving factors not obtained through valid questionnaires from the set of significant influencing factors, and obtaining the final preference results; S32. Using the final preference result data and combining it with a clustering algorithm, a set of classification threshold standards for the levels of influencing factors of the on-the-way service solutions for autonomous driving passenger vehicles is obtained.

7. The method for configuring on-the-way services for autonomous driving passenger vehicles based on user portraits according to claim 6, characterized in that: The S31 comprises the following steps: S311, sorting out the preference result data of users for the factors with significant influencing factors in the valid questionnaires to obtain the questionnaire preference result data; collecting the satisfaction score data of users under the changes of various driving factors through the driving simulator to obtain the driving factor preference result data; the driving factors include the three-axis acceleration change, speed, and driving stability; S312: Merge the questionnaire preference results with the driving factor preference results to obtain final preference result data.

8. The method for configuring on-the-way services for autonomous driving passenger vehicles based on user portraits according to claim 6, characterized in that: The S32 comprises the following steps: S321, use clustering algorithm to perform clustering analysis on the final preference result data, and obtain the optimal clustering number set d = {d1, d2, ..., d i ,...d m }; where d i represents the optimal cluster number of the influencing factors of the i-th autonomous driving passenger car on-the-way service plan; the optimal cluster number represents the number of riding preferences corresponding to each influencing factor in the set of significant influencing factors; S322. Each cluster in the optimal clustering number set represents a group of samples with similar riding preference characteristics. The distance from the sample in each cluster to the center point of the cluster to which it belongs is calculated to obtain the classification threshold standard set e = {e1, e2, ..., e i ,...e m }; where e i Indicates the thresholds corresponding to various riding preferences for each influencing factor.

9. The method for configuring on-the-way services for autonomous driving passenger vehicles based on user portraits according to claim 1, characterized in that: The S4 comprises the following steps: S41. Based on the classification threshold standard set, characterize the user portrait of the autonomous driving passenger car in the valid questionnaire to obtain the optimal number of classifications; S42. Analyze the user group characteristics of each category and the corresponding on-the-way service plan results under the optimal number of categories, and store the user group characteristics and on-the-way service plans of each category; after the users of this category place an order for the autonomous driving passenger car, configure the on-the-way service plan based on historical data.

10. The method for configuring on-the-way services for autonomous driving passenger vehicles based on user portraits according to claim 9, characterized in that: The S41 comprises the following steps: S411, sorting out the basic information and travel habit data in the aforementioned valid questionnaires, and converting the questionnaire data into structured feature vector data in combination with a classification threshold standard set; S412. Using latent profile analysis in Mplus software to classify user profiles of data in structured feature vector data; Set the number of categories k = 1, and gradually increase the value of k, and obtain the optimal number of categories by observing AIC, BIC, ABIC, entropy, likelihood ratio test index LMR, and Bootstrap-based likelihood ratio test index BLRT.

Images