Method, device and electronic device for determining automobile R&D index based on driving behavior

Abstract

The present application relates to a method, device and electronic device for determining automobile R&D indicators based on driving behavior, including obtaining user driving behavior data and VOC data, analyzing the user driving behavior data, dividing the driving condition characteristic area, analyzing the VOC data, extracting the user's attention index items, constructing a KANO model for automobile R&D based on the driving condition characteristic area and the user's attention index items, obtaining VOC index items based on the KANO model, analyzing the VOC index items based on the QFD model, determining the index item test plan, analyzing the correlation coefficient of the test data of the index item test plan, determining the correlation factors that meet the correlation coefficient screening conditions, combining the correlation factors and statistically analyzing, extracting the correlation factor combination that is less than the probability threshold and greater than the determination coefficient, and determining the automobile R&D indicators. Through this method, it is possible to determine the user's highly perceived target items based on driving behavior and VOC data and use them as the basis for automobile R&D.

Description

Technical Field

[0001] The present application relates to the field of automobile product research and development, and in particular to a method, device and electronic equipment for determining automobile research and development indicators based on driving behavior. Background Art

[0002] Automotive product development has always attached great importance to the voice of the customer (VOC). Through market analysis, we can obtain a portrait of potential consumers, and then prepare relatively appropriate, closed-ended survey questionnaires for these consumers to grasp market trends and consumer needs, and provide scientific decision-making basis for automakers to develop new models.

[0003] However, the feedback descriptions of VOC are more life-like. R&D engineers are eager to get in-depth feedback and take into account the balance between automobile systems. The unequal cognition of the two parties often brings confusion to R&D engineers. In automobile R&D, it is necessary to simulate the market feedback of VOC through the engineering language of subjective evaluation, and use this as the evaluation standard for the next stage of objective development. However, there are the following problems. First, due to the lack of derivation basis for the conversion of VOC into subjective evaluation indicators, the subjective evaluation indicators lack VOC information from the market; second, the evaluation standards of subjective evaluation themselves lack standard and weight support, resulting in distortion of the evaluation results themselves; third, when subjective evaluation is implemented in physical development, there is also a lack of derivation basis, so that the objective test results cannot correspond to the subjective evaluation results, and the analysis methods of multi-parameter combinations cannot be refined. Therefore, a method is needed to convert the voice of the customer into an objective experiment, determine the feasibility of the R&D method, and use it as the basis for automobile product development. Summary of the invention

[0004] In view of this, the present application proposes a method for determining automobile R&D indicators based on driving behavior to solve the problems reflected in the above-mentioned background technology.

[0005] According to one aspect of the present application, a method for determining an automobile R&D index based on driving behavior is provided, comprising:

[0006] Obtain user driving behavior data and VOC data;

[0007] Analyzing the user's driving behavior data and dividing the driving condition characteristic areas;

[0008] Analyze the VOC data and extract user-focused index items;

[0009] According to the driving condition characteristic area and the user concerned index item, a KANO model for automobile development is constructed, and a VOC index item is obtained based on the KANO model;

[0010] Analyze the VOC index items based on the QFD model and determine the index item test plan;

[0011] Analyze the correlation coefficient of the test data of the indicator item test plan to determine the relevant factors that meet the correlation coefficient screening conditions;

[0012] The related factors are combined and statistically analyzed to extract the combination of related factors that are less than the probability threshold and greater than the determination coefficient, and determine the automobile research and development index.

[0013] As an optional implementation scheme of the present application, analyzing the user driving behavior data and dividing the driving condition characteristic areas includes:

[0014] Collecting user driving scene data, classifying the user driving scene data, and obtaining driving scene classification data;

[0015] Analyzing the user driving behavior data according to the driving scene classification data to identify the user driving behavior;

[0016] The driving condition characteristic areas are divided according to the driving scene classification data and the driving behavior.

[0017] As an optional implementation scheme of the present application, analyzing the VOC data and extracting the user's attention index items include:

[0018] Cluster analyzing the VOC data to obtain VOC data categories;

[0019] The mention rate of the VOC data category is counted, and the user attention index item is extracted from the corresponding VOC data category according to the mention rate.

[0020] As an optional implementation scheme of the present application, the VOC index items are analyzed based on the QFD model to determine the index item test plan, including:

[0021] Extract the VOC index item and set it as the first left wall and the first roof of the QFD model;

[0022] In the correlation matrix of the QFD model, the correlation of the VOC index items of the first left wall and the first roof is scored to obtain a correlation score;

[0023] Calculating the importance of each VOC indicator item of the first roof, and calculating an average value of the importance proportions according to the importance;

[0024] Based on the QFD model, the VOC index items exceeding the average value are analyzed to determine the index item test plan.

[0025] As an optional implementation scheme of the present application, based on the QFD model, the VOC index items exceeding the average value are analyzed to determine the index item test plan, including:

[0026] The VOC index item exceeding the average value is set as the second left wall of the QFD model, and the objective feature catalog index item is set as the second roof of the QFD model;

[0027] In the correlation matrix of the QFD model, the VOC index items that respectively correspond to the second left wall and the second roof and exceed the average value when they are correlated are counted to determine an index item test plan.

[0028] As an optional implementation scheme of the present application, before analyzing the correlation of the test data of the indicator item test scheme and determining the correlation factors that meet the correlation screening conditions, the method further includes:

[0029] Preset the weight coefficient of the scorer;

[0030] The VOC index item of the first left wall is scored, and a test score and a mean value of the VOC index item are obtained by weighted calculation according to the weight coefficient.

[0031] As an optional implementation scheme of the present application, the correlation coefficient of the test data of the indicator item test scheme is analyzed to determine the relevant factors that meet the correlation coefficient screening conditions, including:

[0032] Segmenting and processing the test data of the indicator item test scheme to obtain the test data of the objective feature catalog indicator item;

[0033] Calculate the characteristic value and mean value of the test data of the objective characteristic catalog index item;

[0034] Analyze the correlation coefficient between the evaluation score, the mean value of the VOC index item, and the characteristic value and mean value of the test data of the objective feature catalog index item;

[0035] When the correlation coefficient meets the correlation coefficient screening condition, it is determined that the VOC index item of the first roof and the characteristic catalog index item of the second roof are correlation factors.

[0036] The present application also provides a device for determining automobile R&D indicators based on driving behavior, comprising:

[0037] The data acquisition module is used to obtain user driving behavior data and VOC data;

[0038] A module for dividing driving condition characteristic regions, used for analyzing the user driving behavior data and dividing driving condition characteristic regions;

[0039] A module for extracting attention index items, used to analyze the VOC data and extract user attention index items;

[0040] A VOC index item acquisition module is used to construct a KANO model for automobile development according to the driving condition characteristic area and the user-focused index item, and obtain the VOC index item based on the KANO model;

[0041] An indicator item test plan acquisition module is used to analyze the VOC indicator item based on a QFD model and determine an indicator item test plan;

[0042] Obtaining a correlation factor module, used to analyze the correlation coefficient of the test data of the indicator item test scheme, and determine the correlation factors that meet the correlation coefficient screening conditions;

[0043] The automobile R&D index module is obtained to combine the related factors and perform statistical analysis, extract the combination of related factors that is less than the probability threshold and greater than the determination coefficient, and determine the automobile R&D index.

[0044] As an optional implementation scheme of the present application, the module for dividing driving condition characteristic regions includes:

[0045] A scene classification data acquisition module is used to collect user driving scene data, classify the user driving scene data, and obtain driving scene classification data;

[0046] A driving behavior identification module is used to classify the data according to the driving scenario, analyze the user driving behavior data, and identify the user driving behavior;

[0047] The data analysis module is used to divide the driving condition characteristic area according to the driving scene classification data and the driving behavior.

[0048] The present application also provides an electronic device, including:

[0049] processor;

[0050] a memory for storing processor-executable instructions;

[0051] Wherein, the processor is configured to implement the above-mentioned method for determining automobile R&D indicators based on driving behavior when executing the executable instructions.

[0052] Beneficial effects of this application:

[0053] The present invention acquires user driving behavior data and VOC data, analyzes the user driving behavior data, divides driving condition characteristic areas, analyzes the VOC data, extracts user-focused index items, constructs a KANO model for automobile research and development according to the driving condition characteristic areas and the user-focused index items, obtains VOC index items based on the KANO model, analyzes the VOC index items based on a QFD model, determines an index item test plan, analyzes the correlation coefficient of the test data of the index item test plan, determines the correlation factors that meet the correlation coefficient screening conditions, combines the correlation factors and performs statistical analysis, extracts a combination of correlation factors that is less than a probability threshold and greater than a determination coefficient, determines automobile research and development indicators, and can determine user-highly perceived target items based on driving behavior and VOC data and use them as the basis for automobile research and development, thereby solving many pain points in automobile forward research and development, such as the non-correspondence between engineering index test conditions and user high-frequency driving conditions, certain deviations between user-highly focused demand points in daily driving and core engineering indicators, and unclear correspondence between subjective evaluation indicators and engineering indicators.

[0054] Other features and aspects of the present application will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the present application and, together with the description, serve to explain the principles of the present application.

[0056] Figure 1 A flow chart showing a method for determining automobile R&D indicators based on driving behavior according to an embodiment of the present application;

[0057] Figure 2 A KANO model diagram showing a method for determining automobile R&D indicators based on driving behavior according to an embodiment of the present application;

[0058] Figure 3 A diagram showing a method for determining an automobile development index based on driving behavior according to an embodiment of the present application;

[0059] Figure 4 A driving condition characteristic area diagram showing a method for determining automobile R&D indicators based on driving behavior according to an embodiment of the present application;

[0060] Figure 5 A VOC data category diagram showing a method for determining automobile R&D indicators based on driving behavior according to an embodiment of the present application;

[0061] Figure 6 A block diagram of a device for determining automobile R&D indicators based on driving behavior according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0062] Various exemplary embodiments, features and aspects of the present application will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0063] Among them, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application or simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0064] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0065] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0066] In addition, in order to better illustrate the present application, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present application can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present application.

[0067] In the field of product development and management, VOC is the abbreviation of Voice of Customer, which means listening to the feedback from customers on products. This is a customer-centric R&D concept that emphasizes improving products based on customer feedback to improve product quality and customer satisfaction.

[0068] The KANO model is a quality management tool used to understand and classify customer needs. When applying the KANO model, it is usually necessary to conduct user surveys to collect user evaluations of products or services. The survey questionnaire is generally in the form of a matrix scale, allowing users to positively and negatively evaluate the functions. Then, the data is processed and analyzed based on the evaluation results to determine the type of each need. Through the analysis of the KANO model, companies can have a clearer understanding of the importance and urgency of customer needs, so as to optimize product design and services in a targeted manner. This can not only improve customer satisfaction and loyalty, but also help companies stand out in the fierce market competition.

[0069] The QFD model, namely the Quality Function Deployment model, allows enterprises to more clearly understand customer needs and expectations and transform them into specific design and production requirements. This helps reduce the blindness and risks of product development and improve the market competitiveness of enterprises.

[0070] Example 1

[0071] Figure 1 A flow chart of a method for determining automobile R&D indicators based on driving behavior according to an embodiment of the present application is shown. Through this method, after processing VOC, the basis for vehicle model development can be determined. Figure 1 As shown, the flow chart includes:

[0072] S100, obtaining user driving behavior data and VOC data;

[0073] Driving behavior data is obtained through user provision, and VOC data is obtained by crawling massive VOC data of competing vehicles through web crawlers.

[0074] S200, analyzing the user's driving behavior data and dividing the driving condition characteristic areas;

[0075] Through driving behavior analysis, a set of user high-frequency driving behavior analysis methods has been established to divide the driving condition characteristic areas and distinguish the definitions of comfort zone, emergency zone, and limit zone. Among them, user driving behavior data refers to high-frequency driving behavior data, and high-frequency driving behavior refers to the behavior of driving multiple times in a short period of time. Specifically, this may include frequent driving actions such as starting, accelerating, braking, turning, etc. in a short period of time, or frequent driving in specific environments such as peak hours and congested roads.

[0076] S300, analyzing the VOC data to extract user-focused index items;

[0077] By analyzing and screening the VOC data, we can obtain the indicators that users pay high attention to. The indicators that users pay high attention to are the user-focused indicators and are used for the subsequent construction of the KANO model.

[0078] S400, constructing a KANO model for automobile development according to the driving condition characteristic area and the user-focused index item, and obtaining a VOC index item based on the KANO model;

[0079] like Figure 2 , combining the driving behavior analysis of S200 and the customer VOC analysis results of S300, constructing the KANO model through customer cognitive coverage and the performance boundary that customers can perceive, and taking the target items that users highly perceive as VOC indicators. The target items that users highly perceive correspond to the high attention of the customer VOC analysis part in the figure and the high perception part in the KANO model.

[0080] S500, analyzing the VOC index items based on the QFD model and determining the index item test plan;

[0081] VOE is the abbreviation of Voice of Employee, which means listening to the opinions and suggestions from employees on product development. The unique insights and creativity of employees on product development are the key to promoting product progress. Therefore, listening to VOE is also an indispensable part of the product development process. Using the QFD model, the user's highly perceived target items are transformed into VOE, which can make the description language of VOC lifelike and not concerned about technical details into deep feedback to eliminate superficial problems. It can determine whether the subjective index items of VOC are related to a certain objective test in the development of automotive products, determine the test plan, and the results of the test plan will affect the development of automotive products.

[0082] S600, analyzing the correlation coefficient of the test data of the indicator item test scheme, and determining the correlation factors that meet the correlation coefficient screening conditions;

[0083] Based on the QFD model analysis, the VOC index items are divided into two stages. The first stage of QFD is the transformation from VOC index items to subjective evaluation index items; the second stage of QFD is the transformation from subjective evaluation index items to objective tests. After determining the subjective evaluation index items obtained in the first stage and the objective index items obtained in the second stage, the subjective evaluation index items are subjectively scored, and the objective index items are objectively tested. The results of the two are combined into the test data of the index item test plan. The correlation coefficient of the subjective score and the objective data of the objective test is calculated by the following formula. The correlation coefficient is greater than 85% as the screening condition. The correlation coefficient can reflect the correlation between the subjective evaluation index items and the objective index items. When the correlation coefficient exceeds 85%, it is a high correlation, and the subjective evaluation index items and the objective index items are strong correlation factors.

[0084]

[0085] S700, combining the related factors and performing statistical analysis to extract the combination of related factors that are less than a probability threshold and greater than a determination coefficient threshold, and determining automobile R&D indicators.

[0086] like Figure 3 , combine multiple groups of strongly correlated factors, use the least squares method to solve the multivariate regression equation, build a subjective and objective model based on the strongly correlated factors, and use the Matlab data statistics toolkit to calculate the P value and R_sq value. The P value is the probability value, and the R_sq value is the determination coefficient. Here, the probability threshold is 0.1, and the determination coefficient threshold is 85%. The P value that meets the design requirements must be less than 0.1 and the R_sq value must be greater than 85%. Therefore, a model with a P value less than 0.1 and an R_sq value greater than 85%, which meets the requirements of engineering theory, is selected as the final output.

[0087] By integrating VOC indicators with high-frequency driving behaviors, the indicators that users pay close attention to are more convincing. Through the verification and analysis of the subjective and objective models, the feasibility of the research and development methods corresponding to multiple groups of strongly correlated factors is determined, and this is used as the basis for publishing high-perception indicators, determining automotive research and development indicators, and using them for vehicle model development.

[0088] As an optional implementation scheme of the present application, in step S200, analyzing the user driving behavior data and dividing the driving condition characteristic areas includes:

[0089] S201, collecting user driving scene data, classifying the user driving scene data, and obtaining driving scene classification data;

[0090] The high-frequency driving behavior data includes the user's high-frequency driving scene data. The user's high-frequency driving scene data is classified into scenes. The classification dimensions include regular scenes such as cities, highways, passing traffic lights, and on and off ramps to obtain driving scene classification data.

[0091] The classification dimensions of users' high-frequency driving scenario data can also be divided into three categories: commuting, non-commuting travel on weekdays, and travel on holidays, mainly involving urban, highway, and smooth road conditions. It can also simulate users' commuting scenarios for route planning, including: emergency braking at intersections, sudden acceleration, emergency braking at high speeds, high-speed overtaking, turning at intersections, turning at ramps, damaged road surfaces, and jointed road surfaces. R&D personnel can determine the classification dimensions and simulated routes based on their needs to divide the driving condition characteristic areas.

[0092] S202, analyzing the user driving behavior data according to the driving scene classification data, and identifying the user driving behavior;

[0093] After determining the driving scenario classification data, we use frequency analysis, normal distribution, skewed distribution, linear regression and other methods to analyze the user driving behavior data in the classified scenarios to identify the user's actual vehicle braking, steering, acceleration, and driving behaviors such as going over potholes and bumps.

[0094] S203: Divide the driving condition characteristic areas according to the driving scene classification data and the driving behavior.

[0095] Combining driving scenarios and driving behaviors, three driving condition characteristic areas are divided, such as Figure 4 , the characteristic areas of driving conditions include comfort zone, emergency zone and limit zone. The 99% state part of the customer's driving condition is defined as the comfort zone, the remaining 1% is the emergency zone, and the area beyond the user's daily driving is the limit zone. More specifically, the high-frequency area corresponding to the comfort zone in the figure is a high-frequency driving scenario for users, such as cities and other frequent travel areas. The driving scenario in the emergency zone is relatively low-frequency driving scenario, such as travel areas such as highways, and the frequency is lower than that of high-frequency driving scenarios. The limit zone is a driving scenario that exceeds the user's daily driving scenario.

[0096] By analyzing high-frequency driving behavior data, we can distinguish the definitions of comfort zone, emergency zone and limit zone, so as to enable the research and development of automobiles to perform better when high-frequency driving behaviors are adopted in high-frequency driving scenarios.

[0097] As an optional implementation scheme of the present application, in step S300, the VOC data is analyzed to extract the user's attention index items, including:

[0098] S301, cluster analysis of the VOC data to obtain VOC data categories;

[0099] The process of obtaining VOC data is as follows: select competing models based on the market positioning of the R&D model, use web crawlers to crawl down the massive VOC data of competing cars as a whole, perform cluster analysis on the massive VOC data, and simply classify them according to comfort, space, power, appearance, endurance, etc.

[0100] S302, counting the mention rate of the VOC data category, and extracting user attention index items from the corresponding VOC data category according to the mention rate.

[0101] Summarize the mention rate of each category to obtain the topic category with higher attention, and define the range of high perception indicators in the category label with higher attention. It should be noted that the mention rate can be sorted from large to small, and the category labels corresponding to the top three mention rates are selected as the topic category with higher attention. The scope of determination of higher attention is determined by R&D engineers according to the situation and is not limited here.

[0102] Each topic category has several user attention indicators. The mention rate of each user attention indicator is counted, and the user attention indicator items with a mention rate greater than 2% are screened out and defined as the range of high perception indicators. The user attention indicator items with a mention rate greater than 2% are interpreted semantically, and non-engineered and ambiguous semantics are reconfirmed through in-depth interviews. Figure 5 , the user attention indicators with the classification label of dynamics include: push-back feeling, fast start, fast acceleration, and sufficient high-speed power, etc., among which the mention rate reflects the user's attention. Through semantic interpretation and in-depth interviews of user attention indicators with a mention rate greater than 2%, we identify users' effective VOC data, decompose VOC elements, analyze what users are in what scenarios, what they talk about, and what they feel, so as to extract indicators that users pay high attention to.

[0103] As an optional implementation scheme of the present application, in step S500, the VOC index items are analyzed based on the QFD model to determine the index item test plan, including:

[0104] S501, extracting the VOC index item and setting it as the first left wall and the first roof of the QFD model;

[0105] In the KANO model, the target item with high user perception is used as the VOC indicator item. In this embodiment, the target item with high user perception is defined as the “left wall” of QFD as a i , the modified four-level index of subjective evaluation of dynamics is used as the “roof” and is defined as b j It should be noted that the four-level indicators of dynamic subjective evaluation are included in the target items with high user perception, and a i and b j The corresponding user highly perceived target items are not fixed. The user highly perceived target items determined from the aforementioned KANO model are continuously rotated, and the first left wall and the first roof are determined according to actual needs.

[0106] S502, scoring the correlation of the VOC index items of the first left wall and the first roof in the correlation matrix of the QFD model to obtain a correlation score;

[0107] In the correlation matrix part in the middle of the QFD model, for any item a i , subjective evaluation correlation b j The lower part is scored as 1 point, the part with certain correlation but not strong correlation is scored as 3 points, the part with strong correlation is scored as 9 points, and the part with no correlation is left blank, and the default score is 0 points. In general, there are usually no more than 2 strong correlations for a certain user's highly perceived target item, that is, for any one item a i , in the subjective evaluation correlation b j In the test, the number of items with a score of 9 for strong correlation does not exceed two.

[0108] S503, calculating the importance of each VOC indicator item of the first roof, and calculating an average value of the importance proportion according to the importance;

[0109] After completing the relevance scoring, calculate each item b according to the following formula j Project Importance j :

[0110]

[0111] Then, the importance ratio ImpRatio is calculated according to the following formula: j :

[0112]

[0113] Finally, find ImpRatio j The average value of will be greater than the mean b j It is a high-perception subjective indicator item.

[0114] S504, based on the QFD model, analyzing the VOC index items exceeding the average value, and determining an index item test plan.

[0115] After determining the highly perceived subjective indicator items, they are further analyzed based on the QFD model, and finally the indicator item test plan is determined.

[0116] As an optional implementation scheme of the present application, based on the QFD model, the VOC index items exceeding the average value are analyzed to determine the index item test plan, including:

[0117] The VOC index item exceeding the average value is set as the second left wall of the QFD model, and the objective feature catalog index item is set as the second roof of the QFD model;

[0118] In the second stage, the existing objective feature catalog was adjusted by referring to the feature catalog, relevant national standards, industry standards, enterprise standards and Ford R202 test standards. This is a more detailed supplement to the traditional objective test feature catalog.

[0119] In the correlation matrix of the QFD model, the VOC index items that respectively correspond to the second left wall and the second roof and exceed the average value when they are correlated are counted to determine an index item test plan.

[0120] b will be greater than the mean jThe QFD model is constructed by taking the highly perceived subjective indicator as the left wall of the QFD model and the adjusted objective feature catalog as the roof of the QFD model. In the matrix of the QFD model, 1 or 0 is used to determine whether a certain highly perceived subjective indicator is related to a certain objective test, that is, whether it is related to the adjusted objective feature catalog, and the objective feature catalog that is not related to all highly perceived subjective indicator items is eliminated. Finally, the objective feature catalog related to the highly perceived subjective indicator items is determined to form a test plan, that is, the indicator item test plan. The introduction of the technology of the quality house into the engineering transformation of highly perceived subjective indicators can enable engineers to obtain in-depth feedback.

[0121] As an optional implementation scheme of the present application, before analyzing the correlation of the test data of the indicator item test scheme and determining the correlation factors that meet the correlation screening conditions, it also includes:

[0122] Preset the weight coefficient of the scorer;

[0123] For professional engineers who participate in scoring, a weight coefficient needs to be set. Appropriate weight coefficients are given according to personnel positions, professions and levels. The specific weighting rules are determined according to actual conditions and are not limited here.

[0124] The VOC index item of the first left wall is scored, and a test score and a mean value of the VOC index item are obtained by weighted calculation according to the weight coefficient.

[0125] Organize professional engineers to conduct subjective evaluation and scoring of target items with high user perception, and then perform weighted calculation on the subjective evaluation scores to obtain the evaluation scores. Calculate the average of the evaluation scores of multiple professional engineers for a target item with high user perception, where the evaluation score of a target item with high user perception is defined as y i , the mean is expressed as i represents the Nth vehicle.

[0126] As an optional implementation scheme of the present application, the correlation coefficient of the test data of the indicator item test scheme is analyzed to determine the relevant factors that meet the correlation coefficient screening conditions, including:

[0127] Segmenting and processing the test data of the indicator item test scheme to obtain the test data of the objective feature catalog indicator item;

[0128] According to the objective feature catalog related to the high-perception subjective index items, the index item test plan is formed, and the test is carried out according to the index item test plan to obtain the test data of the index item test plan. The test data of the objective feature catalog index items is obtained by conducting objective vehicle tests. That is, when the ImpRatio is greater than jAfter obtaining the high-perceived subjective index item with the average value, determine whether it is related to the adjusted objective feature catalog. If so, the corresponding objective feature catalog is part of the index item test plan. When conducting an objective test, the test data of the corresponding objective feature catalog is obtained. It should be noted that the test data of the index item test plan includes subjective scoring of subjective evaluation index items and objective testing of objective index items. Among them, the subjective scoring of subjective evaluation index items is that professional engineers conduct subjective evaluation and scoring of the user's highly perceived target items, and then weighted calculation of the subjective evaluation scores is performed to obtain the evaluation scores; objective testing of objective indicator items is the test data of objective feature catalog index items.

[0129] Calculate the characteristic value and mean value of the test data of the objective characteristic catalog index item;

[0130] Matlab is used to segment the test data of the objective feature catalog and then solve the eigenvalue. Each eigenvalue is defined as x i , the mean is expressed as i represents the Nth vehicle.

[0131] Analyze the correlation coefficient between the evaluation score, the mean value of the VOC index item, and the characteristic value and mean value of the test data of the objective feature catalog index item;

[0132] The evaluation score and the mean of the VOC index item are for the target items that are highly perceived by users, and the characteristic value and mean of the test data are for the objective feature catalog related to the target items that are highly perceived by users. According to the following formula, the correlation coefficient between each objective feature catalog and the target items that are highly perceived by users can be solved. In addition, the subjective and objective model constructed by the correlation coefficient of the evaluation score, the mean of the VOC index item, and the characteristic value and mean of the test data of the objective feature catalog index item can better reflect the feasibility of automobile research and development.

[0133]

[0134] When the correlation coefficient meets the correlation coefficient screening condition, it is determined that the VOC index item of the first roof and the characteristic catalog index item of the second roof are correlation factors.

[0135] In this embodiment, the correlation factor is a strong correlation factor. The correlation coefficient screening condition is that when the correlation coefficient is greater than 85%, the user's highly perceived target item and its related objective feature catalog are used as strong correlation factors. The range of strongly perceived performance indicators in the user's driving performance is determined through the association analysis of the user's highly perceived target item and its related objective feature catalog.

[0136] Through the above method, it is possible to determine the target items that users highly perceive based on driving behavior and VOC data and use them as the basis for automobile research and development. This can solve many pain points in automobile forward development, such as the mismatch between engineering indicator test conditions and users' high-frequency driving conditions, the certain deviation between users' high-paying demand points in daily driving and core engineering indicators, and the unclear correspondence between subjective evaluation indicators and engineering indicators.

[0137] Example 2

[0138] Based on the same principle as the above method, a device for determining automobile R&D indicators based on driving behavior is also proposed, see Figure 6 , a device 100 for determining automobile R&D indicators based on driving behavior according to an embodiment of the present disclosure includes:

[0139] 110, a data acquisition module, used to acquire user driving behavior data and VOC data;

[0140] 120, a module for dividing driving condition characteristic regions, used for analyzing the user driving behavior data and dividing driving condition characteristic regions;

[0141] 130, a module for extracting attention index items, used to analyze the VOC data and extract the user attention index items;

[0142] 140, a VOC index item obtaining module, which is used to construct a KANO model for automobile research and development according to the driving condition characteristic area and the user-focused index item, and obtain a VOC index item based on the KANO model;

[0143] 150, obtaining an index item test plan module, used to analyze the VOC index item based on a QFD model and determine an index item test plan;

[0144] 160, obtaining a correlation factor module, for analyzing the correlation coefficient of the test data of the indicator item test scheme, and determining the correlation factors that meet the correlation coefficient screening conditions;

[0145] 170, obtaining an automobile R&D index module, for combining the related factors and statistically analyzing them, extracting a combination of related factors that is less than a probability threshold and greater than a determination coefficient, and determining an automobile R&D index.

[0146] As an optional implementation scheme of the present application, the module for dividing driving condition characteristic regions includes:

[0147] 121, obtaining a scene classification data module, used to collect user driving scene data, classify the user driving scene data, and obtain driving scene classification data;

[0148] 122, a driving behavior identification module, configured to analyze the user driving behavior data according to the driving scenario classification data, and identify the user driving behavior;

[0149] 123, a data analysis module, used to divide the driving condition characteristic area according to the driving scene classification data and the driving behavior.

[0150] Obviously, those skilled in the art should understand that all or part of the processes in the above-mentioned embodiment method can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned control methods. The modules or steps of the present invention can be implemented by a general-purpose computing device. They can be concentrated on a single computing device or distributed on a network composed of multiple computing devices. Optionally, they can be implemented by program codes executable by computing devices, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules respectively, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. In this way, the present invention is not limited to any specific combination of hardware and software.

[0151] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment method can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned control methods. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (Flash Memory), a hard disk (Hard Disk Drive, abbreviated as: HDD) or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above-mentioned types of memory.

[0152] Example 3

[0153] Furthermore, an electronic device is proposed, comprising:

[0154] processor;

[0155] a memory for storing processor-executable instructions;

[0156] Wherein, the processor is configured to implement the method for determining automobile R&D indicators based on driving behavior described in Example 1 when executing the executable instructions.

[0157] The electronic device of the embodiment of the present disclosure includes a processor and a memory for storing instructions executable by the processor, wherein the processor is configured to implement any of the above-mentioned methods for determining automobile R&D indicators based on driving behavior when executing the executable instructions.

[0158] It should be noted that the number of processors may be one or more. At the same time, the electronic device of the embodiment of the present disclosure may also include an input device and an output device. The processor, memory, input device, and output device may be connected via a bus or in other ways, which are not specifically limited here.

[0159] The memory is a computer-readable storage medium for the method for determining automobile R&D indicators based on driving behavior, and can be used to store software programs, computer executable programs, and various modules, such as: the program or module corresponding to the method for determining automobile R&D indicators based on driving behavior in the embodiment of the present disclosure. The processor executes various functional applications and data processing of the electronic device by running the software programs or modules stored in the memory.

[0160] The input device can be used to receive input numbers or signals. The signal can be a key signal related to user settings and function control of the device / terminal / server. The output device can include a display device such as a display screen.

[0161] The embodiments of the present application have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for determining automobile R&D indicators based on driving behavior, characterized in that: include: Obtain user driving behavior data and VOC data; Analyzing the user's driving behavior data and dividing the driving condition characteristic areas; Analyze the user's driving behavior data and divide the driving condition characteristic areas, including: Collecting user driving scene data, classifying the user driving scene data, and obtaining driving scene classification data; Analyzing the user driving behavior data according to the driving scene classification data to identify the user driving behavior; Dividing driving condition characteristic areas according to the driving scene classification data and the driving behavior; Analyze the VOC data and extract user-focused index items; Analyze the VOC data and extract the user's attention indicators, including: Cluster analyzing the VOC data to obtain VOC data categories; Counting the mention rate of the VOC data category, and extracting the user attention index item from the corresponding VOC data category according to the mention rate; According to the driving condition characteristic area and the user concerned index item, a KANO model for automobile development is constructed, and a VOC index item is obtained based on the KANO model; Analyze the VOC index items based on the QFD model and determine the index item test plan; The VOC index items are analyzed based on the QFD model, and the index item test plan is determined, including: Extract the VOC index item and set it as the first left wall and the first roof of the QFD model; In the correlation matrix of the QFD model, the correlation of the VOC index items of the first left wall and the first roof is scored to obtain a correlation score; Calculating the importance of each VOC indicator item of the first roof, and calculating an average value of the importance proportions according to the importance; Based on the QFD model, analyzing the VOC index items exceeding the average value, and determining the index item test plan; Based on the QFD model, the VOC index items exceeding the average value are analyzed to determine the index item test plan, including: The VOC index item exceeding the average value is set as the second left wall of the QFD model, and the objective feature catalog index item is set as the second roof of the QFD model; In the correlation matrix of the QFD model, the VOC index items respectively corresponding to the second left wall and the second roof and exceeding the average value when they are correlated are counted to determine the index item test plan; Analyze the correlation coefficient of the test data of the indicator item test plan to determine the relevant factors that meet the correlation coefficient screening conditions; Analyze the correlation coefficient of the test data of the indicator test plan to determine the relevant factors that meet the correlation coefficient screening conditions, including: Segmenting and processing the test data of the indicator item test scheme to obtain the test data of the objective feature catalog indicator item; Calculate the characteristic value and mean value of the test data of the objective characteristic catalog index item; Analyze the correlation coefficient between the evaluation score, the mean value of the VOC index item, and the characteristic value and mean value of the test data of the objective feature catalog index item; When the correlation coefficient meets the correlation coefficient screening condition, determining that the VOC index item of the first roof and the characteristic catalog index item of the second roof are correlation factors; Before analyzing the correlation of the test data of the indicator item test scheme and determining the relevant factors that meet the correlation screening conditions, it also includes: Preset the weight coefficient of the scorer; Scoring the VOC index item of the first left wall, and obtaining the evaluation score and the average value of the VOC index item by weighted calculation according to the weight coefficient; The related factors are combined and statistically analyzed to extract the combination of related factors that are less than the probability threshold and greater than the determination coefficient, and determine the automobile research and development index.

2. A device for determining automobile R&D indicators based on driving behavior, characterized in that: include: The data acquisition module is used to obtain user driving behavior data and VOC data; A module for dividing driving condition characteristic regions, used for analyzing the user driving behavior data and dividing driving condition characteristic regions; The module for dividing driving condition characteristic regions comprises: A scene classification data acquisition module is used to collect user driving scene data, classify the user driving scene data, and obtain driving scene classification data; A driving behavior identification module is used to classify the data according to the driving scenario, analyze the user driving behavior data, and identify the user driving behavior; A data analysis module, used to divide the driving condition characteristic area according to the driving scene classification data and the driving behavior; A module for extracting attention index items, used to analyze the VOC data and extract user attention index items; Analyze the VOC data and extract the user's attention indicators, including: Cluster analyzing the VOC data to obtain VOC data categories; Counting the mention rate of the VOC data category, and extracting the user attention index item from the corresponding VOC data category according to the mention rate; A VOC index item acquisition module is used to construct a KANO model for automobile development according to the driving condition characteristic area and the user-focused index item, and obtain the VOC index item based on the KANO model; An indicator item test plan acquisition module is used to analyze the VOC indicator item based on a QFD model and determine an indicator item test plan; The VOC index items are analyzed based on the QFD model, and the index item test plan is determined, including: Extract the VOC index item and set it as the first left wall and the first roof of the QFD model; In the correlation matrix of the QFD model, the correlation of the VOC index items of the first left wall and the first roof is scored to obtain a correlation score; Calculating the importance of each VOC indicator item of the first roof, and calculating an average value of the importance proportions according to the importance; Based on the QFD model, analyzing the VOC index items exceeding the average value, and determining the index item test plan; Based on the QFD model, the VOC index items exceeding the average value are analyzed to determine the index item test plan, including: The VOC index item exceeding the average value is set as the second left wall of the QFD model, and the objective feature catalog index item is set as the second roof of the QFD model; In the correlation matrix of the QFD model, the VOC index items respectively corresponding to the second left wall and the second roof and exceeding the average value when they are correlated are counted to determine the index item test plan; Obtaining a correlation factor module, used to analyze the correlation coefficient of the test data of the indicator item test scheme, and determine the correlation factors that meet the correlation coefficient screening conditions; Analyze the correlation coefficient of the test data of the indicator test plan to determine the relevant factors that meet the correlation coefficient screening conditions, including: Segmenting and processing the test data of the indicator item test scheme to obtain the test data of the objective feature catalog indicator item; Calculate the characteristic value and mean value of the test data of the objective characteristic catalog index item; Analyze the correlation coefficient between the evaluation score, the mean value of the VOC index item, and the characteristic value and mean value of the test data of the objective feature catalog index item; When the correlation coefficient meets the correlation coefficient screening condition, determining that the VOC index item of the first roof and the characteristic catalog index item of the second roof are correlation factors; Before analyzing the correlation of the test data of the indicator item test scheme and determining the relevant factors that meet the correlation screening conditions, it also includes: Preset the weight coefficient of the scorer; Scoring the VOC index item of the first left wall, and obtaining the evaluation score and the average value of the VOC index item by weighted calculation according to the weight coefficient; The automobile R&D index module is obtained to combine the related factors and perform statistical analysis, extract the combination of related factors that is less than the probability threshold and greater than the determination coefficient, and determine the automobile R&D index.

3. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to implement the method for determining automobile R&D indicators based on driving behavior as described in claim 1 above when executing the executable instructions.

Images