Automobile maintenance price comparison evaluation method based on multi-dimensional data fusion

Abstract

The invention discloses an automobile maintenance price comparison evaluation method based on multi-dimensional data fusion, and particularly relates to the technical field of automobile service, and the method comprises the steps: firstly collecting the price data, time cost data and maintenance quality data of automobile maintenance price comparison, and then constructing a corresponding price, time cost and maintenance quality evaluation model; and performing two-stage adaptive weight adjustment, calculating a multi-dimensional price comparison comprehensive score, presenting a final price comparison result to an automobile user client, performing deviation analysis on a maintenance merchant result selected by an automobile user, and performing automobile maintenance price comparison evaluation process optimization according to a deviation analysis result. According to the invention, a price comparison price evaluation model, a price comparison time cost evaluation model and a price comparison maintenance quality evaluation model are constructed; and meanwhile, a multi-dimensional price comparison comprehensive score is calculated through model adaptive weight basic optimization and deep optimization, so that an evaluation result is more scientific and reliable, and a vehicle owner is helped to make a judicious decision based on data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automobile service, in particular to a vehicle maintenance and repair price evaluation method based on multi-dimensional data fusion. BACKGROUND

[0002] With the continuous growth of the number of cars, the market size of vehicle maintenance and repair is expanding, and when choosing maintenance and repair services, car owners often face the problems of information asymmetry and single evaluation method. For car owners, how to compare prices among many vehicle maintenance and repair service providers and make a wise choice based on the comparison and evaluation results is a key problem.

[0003] The existing price comparison system mostly only makes static comparison based on the price of maintenance and repair projects, provides price information of each repair shop for different projects, and ignores the influence of time cost, maintenance quality, maintenance price and other key factors on the overall service value. It almost does not involve the evaluation of maintenance and repair time cost, such as user travel time consumption, business response timeliness, and does not involve the consideration of maintenance quality, including the professional level of maintenance technicians, the quality of spare parts, and the standardization of maintenance technology.

[0004] The current vehicle maintenance and repair market has many deficiencies in price comparison and evaluation, for example: the existing price comparison system only evaluates the price, generally uses a static evaluation model, that is, the data is not updated in time, cannot reflect the dynamic fluctuations of market changes and service quality of repair shops in real time, does not establish a detailed price comparison price, time cost and maintenance quality quantitative evaluation model, resulting in deviation between the evaluation results and the real experience of users, and further leading to user selection deviation. Therefore, there is an urgent need for a dynamic price comparison and evaluation method that can consider multiple factors to meet the growing demand of car owners for high-quality and efficient maintenance and repair services. SUMMARY

[0005] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present application provide a vehicle maintenance and repair price evaluation method based on multi-dimensional data fusion to solve the problems raised in the background art.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical scheme: a vehicle maintenance and repair price evaluation method based on multi-dimensional data fusion, comprising: S1: collecting price data, time cost data and maintenance quality data of vehicle maintenance and repair price through a data acquisition interface; S2: analyzing the collected price data, time cost data and maintenance quality data respectively, and constructing a price comparison price evaluation model, a price comparison time cost evaluation model and a price comparison maintenance quality evaluation model; S3: Perform two-stage adaptive weight adjustment on the price comparison price evaluation model, the price comparison time cost evaluation model and the price comparison maintenance quality evaluation model to obtain the adaptive weight adjustment results of each model after optimization; S4: Based on the adaptive weight adjustment results, perform fusion calculation of the price comparison price evaluation model, the price comparison time cost evaluation model and the price comparison maintenance quality evaluation model to obtain a multi-dimensional price comparison comprehensive score, correct the multi-dimensional price comparison comprehensive score to generate a price comparison result with a confidence interval, and present the final price comparison result to the automobile user client; S5: Record the maintenance merchant result selected by the automobile user client according to the final price comparison result, perform deviation analysis on the selected maintenance merchant result, and optimize the automobile maintenance price comparison evaluation process according to the deviation analysis result.

[0007] The technical effects and advantages of the present application are as follows: 1. The present application collects price data, time cost data and maintenance quality data of automobile maintenance price comparison through a data collection interface, realizes multi-dimensional factor analysis, changes the traditional static price comparison method, realizes comprehensive and dynamic evaluation of automobile maintenance service in subsequent price comparison evaluation steps, and provides more accurate service value reference for the car owner. 2. The present application constructs a price comparison price evaluation model, a price comparison time cost evaluation model and a price comparison maintenance quality evaluation model through the multi-dimensional price comparison evaluation model calculation step; at the same time, the model adaptive weight basic optimization and deep optimization are used to calculate the multi-dimensional price comparison comprehensive score, so that the evaluation result is more scientific and reliable, and the car owner can make a wise decision based on data. 3. The present application performs deviation analysis on the selected maintenance merchant result through the automobile maintenance price comparison evaluation feedback optimization step, optimizes the automobile maintenance price comparison evaluation process according to the deviation analysis result, provides the latest price comparison information for the car owner, can meet the individualized needs of different car owners at different periods, and improves the car owner's satisfaction. BRIEF DESCRIPTION OF DRAWINGS

[0008] Figure 1 The present application is a whole process schematic diagram.

[0009] Figure 2 The present application is a method flowchart. DETAILED DESCRIPTION

[0010] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0011] Please see Figure 1 As shown, the present invention provides a car repair and maintenance price comparison evaluation system based on multi-dimensional data fusion, including a multi-dimensional price comparison data acquisition module, a multi-dimensional price comparison evaluation model calculation module, a multi-dimensional evaluation model adaptive weight adjustment module, a car repair and maintenance price comparison evaluation result presentation module, and a car repair and maintenance price comparison evaluation feedback optimization module.

[0012] The multidimensional price comparison data acquisition module is connected to the multidimensional price comparison evaluation model calculation module. The multidimensional evaluation model adaptive weight adjustment module is connected to both the multidimensional price comparison evaluation model calculation module and the automobile repair and maintenance price comparison evaluation result presentation module. The automobile repair and maintenance price comparison evaluation feedback optimization module is connected to the automobile repair and maintenance price comparison evaluation result presentation module.

[0013] Multi-dimensional price comparison data acquisition module: Through the data acquisition interface, it collects price data, time cost data, and repair quality data for car repair and maintenance, and transmits the collected data to the multi-dimensional price comparison evaluation model calculation module; Multidimensional price comparison evaluation model calculation module: Analyzes the collected price data, time cost data and repair quality data respectively, constructs a price comparison evaluation model, a time comparison evaluation model and a repair quality comparison evaluation model, and transmits the constructed multidimensional price comparison evaluation model to the multidimensional evaluation model adaptive weight adjustment module; The multidimensional evaluation model adaptive weight adjustment module performs two-stage adaptive weight adjustment on the price comparison evaluation model, the time cost comparison evaluation model, and the repair quality comparison evaluation model, and transmits the adaptive weight adjustment results of each model to the automobile repair and maintenance price comparison evaluation result presentation module. The vehicle repair and maintenance price comparison evaluation result presentation module: Based on the adaptive weight adjustment result, it integrates the price comparison evaluation model, the time cost comparison evaluation model, and the repair quality comparison evaluation model to calculate a multi-dimensional price comparison comprehensive score. Based on the multi-dimensional price comparison comprehensive score, it generates a price comparison result with confidence interval and presents the final price comparison result to the car user's client. The car repair and maintenance price comparison evaluation feedback optimization module records the repair shop selected by the car user client based on the final price comparison results, performs deviation analysis on the selected repair shop results, and optimizes the car repair and maintenance price comparison evaluation process based on the deviation analysis results.

[0014] Please see Figure 2As shown, a method for comparing car repair and maintenance prices based on multi-dimensional data fusion includes: S1: Collecting price data, time cost data, and repair quality data for car repair and maintenance through a data acquisition interface; S2: Analyzing the collected price data, time cost data, and repair quality data respectively, and constructing a price comparison evaluation model, a time comparison evaluation model, and a repair quality comparison evaluation model; S3: Performing a two-stage adaptive weight adjustment on the price comparison evaluation model, the time comparison evaluation model, and the repair quality comparison evaluation model to obtain the optimized adaptive weight adjustment results for each model; S4: Based on the adaptive weight adjustment results, fusing the price comparison evaluation model, the time comparison evaluation model, and the repair quality comparison evaluation model to calculate a multi-dimensional price comparison comprehensive score, correcting the multi-dimensional price comparison comprehensive score to generate a price comparison result with a confidence interval, and presenting the final price comparison result to the car user's client; S5: Recording the repair shop selected by the car user's client based on the final price comparison result, performing a deviation analysis on the selected repair shop result, and optimizing the car repair and maintenance price comparison evaluation process based on the deviation analysis results.

[0015] S1: Collect price data, time cost data, and repair quality data for automotive repair and maintenance through a data acquisition interface; the price data includes the price of the same maintenance item, the market average price, the price of the required replacement parts, the corresponding market average price of the required replacement parts, the labor cost in the maintenance item, and the industry benchmark labor cost; the time cost data includes the repair shop's response time, the number of orders completed on time, and the one-way driving time; the repair quality data includes the number of industry certifications held by technicians, the total number of technicians, the types of parts required for replacement, the quantity of parts from the parts source, the execution record of process steps, and standard maintenance process specifications; This embodiment requires specific explanation of the following: Price data collection: Real-time collection of price information from various repair shops for different car brands, models, and specific maintenance items. This includes not only basic maintenance items (such as oil changes and filter replacements) but also prices for various complex repair items (such as engine overhauls and transmission repairs); Time cost data collection: Obtaining the estimated and actual completion times of different maintenance items from repair shops. Through the repair shop's work order system and car owner feedback, the time from sending the item for repair to delivery is statistically analyzed, and the time consumption of each stage of the repair process (such as diagnosis, parts preparation, repair operation, and quality inspection) is analyzed; Quality data collection: Collecting the quality information of the parts used by the repair shops, including the brand of the parts and whether they are genuine original parts. Analyzing the repair shop's repair quality records, such as return rate and first-time success rate of fault resolution.

[0016] S2: Analyze the collected price data, time cost data, and repair quality data respectively, and construct a price comparison evaluation model, a time cost comparison evaluation model, and a repair quality comparison evaluation model, including the following steps: S2.1: Within the preset evaluation period, through the automotive repair and maintenance industry data platform, collect n1 times the price p collected for the j-th repair shop for the same maintenance item (i.e., repair and maintenance item) for the i-th time. j,i To obtain the market average price fluctuation coefficient η(p) for the same maintenance and repair project, ,μ(p i The price pp represents the average market price at the i-th time of data collection for this maintenance and repair item. The average market price is calculated by averaging the prices of registered maintenance and repair businesses with the same maintenance and repair item from the data platform. Then, the price pp of the j-th maintenance and repair business for the i-th time of data collection is collected n2 times. j,i The average market price μ(pp) corresponding to the required replacement parts i ), thus obtaining the component price fluctuation coefficient η(pp), The average market price of the parts refers to the average price of the parts required for maintenance and repair projects, obtained by filtering out registered repair shops from the automotive repair and maintenance industry data platform and selecting the shops that offer the parts required for maintenance and repair projects, along with their corresponding prices. Then, the j-th repair shop's labor cost (moy) is collected n3 times for the i-th time in the maintenance and repair project. j,i Compared to industry benchmark hourly rates j,0 The labor cost difference coefficient η(moy), Finally, the price comparison evaluation model P1 is obtained, P1=exp-[η(p)+η(pp)+η(moy)]; S2.2: Within the preset evaluation period, collect the time (in minutes) t from the user initiating the maintenance project request to the maintenance provider's first response during the i-th data collection for the j-th maintenance provider (n4 times). i After taking the logarithm and standardizing, the response time index η(t) is obtained. Simultaneously, the number of orders (nd) completed on time by the repair shop within the promised construction time for the n4 maintenance projects is calculated to obtain the construction completion rate η(nd), where η(nd) = nd / n4; The one-way driving time (tg) from the car user to the repair shop in the n4 maintenance projects is also calculated. i (minutes), to obtain the service efficiency coefficient η(tg), Finally, the price comparison time cost assessment model P2 is obtained, P2=η(t)+η(nd)+η(tg); S2.3: Within a preset evaluation period, collect the ratio of the number nh of industry certifications (such as ASE for automotive repair, I-CAR, etc.) held by the technicians of the j-th repair shop to the total number of technicians T_nh, to obtain the technical certification coverage rate η(nh), η(nh) = nh / T_nh; collect the set CT of the source types of replacement parts required for the maintenance project of the j-th repair shop n2 times, CT = [ct1, ct2, ..., ctm], where m is the number of source types of replacement parts, and ctm is the source of the m-th part. The source of parts can be, for example, original parts, brand parts, aftermarket parts, etc., and then count the set n_CT of the number of parts from the m-th part source, n_CT = [n_ct1, n_ct2, ..., n_ctm], where n_ctm is the number of parts from the m-th part source, to obtain the part compatibility rate η(CT). Where n_ctJ represents the quantity of original parts and n_ctp represents the quantity of branded parts; through the repair shop's digital management system, the execution record of the i-th process step during n5 complete maintenance processes is collected and compared with the standard maintenance process specification to obtain the maintenance process specification coefficient η(DP). I() is an exponential function, where I=1 indicates that the maintenance process conforms to the specifications, otherwise I=0, M is the total number of standard maintenance process steps, and q N and q N,0 These are the execution records of the Nth process step and the corresponding standard process step execution records, respectively; finally, the price comparison maintenance quality assessment model P3 is obtained, P3=η(nh)+η(CT)+η(DP); In this embodiment, it should be specifically noted that the conformity judgment of the indicator function I is automatically executed by the rule engine module deployed on the server. In the standard process database, the following information is preset for each process step N of each maintenance project: 1.1 Standard parameter value q N,0 ; 1.2 Parameter data type (e.g., text, numeric, boolean); 1.3. Compliance with standard judgment rules; for example: for discrete text parameters (such as part model, operation name), it is judged to comply with the standard if and only if the string value is exactly the same as the parameter (I=1), otherwise it does not comply (I=0); for continuous numerical parameters (such as torque, pressure, usage), an allowable deviation threshold Δq is set. N When |q N -q N,0 |≤Δq N If the condition is met, it is determined to be compliant (I=1); otherwise, it is not compliant (I=0). For example, for the process sequence parameter, the standard process q... N,0 Defined as an ordered list, when the actual execution sequence q NIf the elements and order of the given list are exactly the same, it is considered compliant (I=1); otherwise, it is not compliant (I=0). For example, for existence parameters (such as quality inspection confirmation), when the standard requires q... N,0 The value is "true" or "yes", and q is actually recorded. N If the result is "true" or "yes", it is considered compliant (I=1); otherwise, it is not compliant (I=0).

[0017] S3: Perform two-stage adaptive weight adjustment on the price comparison evaluation model, the time cost comparison evaluation model, and the repair quality comparison evaluation model to obtain the optimized adaptive weight adjustment results for each model, including the following steps: This embodiment needs to specifically explain that the two-stage adaptive weight adjustment includes a first-stage adaptive weight adjustment and a second-stage adaptive weight adjustment.

[0018] S3.1: First-stage adaptive weight adjustment: Within the preset evaluation period, the basic weight W of the I-th model is dynamically adjusted based on the historical data of each model value (historical data includes historical average, historical maximum, and historical minimum). I , W I，0 P represents the base weights of the i-th model, α is the weight adjustment coefficient, and its value is [0,1]. I (t), P I ,avg、P I ,max and P I ,min represent the current value, historical average value, historical maximum value, and historical minimum value of the I-th model, respectively. I=1 represents the price comparison evaluation model, I=2 represents the price comparison time cost evaluation model, and I=3 represents the price comparison repair quality evaluation model. The basic weights are obtained from historical data using a linear regression model to determine the basic weights of the price comparison evaluation model, the price comparison time cost evaluation model, and the price comparison repair quality evaluation model. S3.2: Second-stage adaptive weight adjustment: Within the preset evaluation period, based on the maintenance requests initiated by car users, the proportion of orders in the car user's historical orders that selected the lowest price option is retrieved as the preference strength Q1 of the price comparison evaluation model. n_dj represents the number of orders that selected the lowest price option, Tnd represents the total number of historical orders, and the percentage of car users who selected original parts is also retrieved as the preference strength Q3 of the price comparison repair quality assessment model. n_ctJ represents the quantity of original parts. Simultaneously, using a natural language processing model, the proportion of orders placed when car users inquired through the client-side interactive platform is obtained, serving as the preference strength Q2 for the price comparison time cost assessment model. Let n_tim be the number of orders at the query time, and obtain the I-th model weight W for deep optimization.I use , Q I Let β be the preference strength of the I-th model obtained based on the analysis of historical behavior of car users. I Let be the preference coefficient of the i-th model. W I，max The preset upper limit for the weight of the i-th model is used to prevent a single indicator from monopolizing the evaluation; finally, the adaptive weight adjustment results of each model after optimization are obtained, including the weight W1 of the price comparison evaluation model. use Price comparison time cost assessment model weight W2 use Price comparison repair quality assessment model weight W3 use ; In this embodiment, it is necessary to specifically explain that firstly, historical data is used to filter out semantic texts related to time, such as "urgent," "as soon as possible," and "what time," which are used as the training set for a natural language processing model (such as the BERT model). The trained natural language processing model is then output to identify orders that inquire about time on the client interaction platform and obtain the number of orders that inquire about time. The natural language processing model is an existing technology.

[0019] S4: Based on the adaptive weight adjustment results, a multi-dimensional price comparison comprehensive score is calculated by fusing the price evaluation model, the time cost evaluation model, and the repair quality evaluation model. The multi-dimensional price comparison comprehensive score is then corrected to generate a price comparison result with confidence intervals, and the final price comparison result is presented to the car user's client. This includes the following steps: S4.1: Within the preset evaluation period, based on the adaptive weight adjustment results, the multi-dimensional price comparison comprehensive score S of the j-th repair shop is calculated by integrating the price comparison evaluation model, the time cost comparison evaluation model, and the repair quality comparison evaluation model. j , P I For the i-th evaluation model value, W I use The I-th model weight is the result of deep optimization. S4.2: Within the preset evaluation period, based on the number of service orders for the j-th repair merchant, adjust the multi-dimensional price comparison comprehensive score to generate the price comparison result S for the j-th repair merchant with a confidence interval. j u , N j 、N0、N min and N max These represent the number of service orders for the j-th repair shop, the industry benchmark number of service orders, the minimum number of orders in the industry, and the maximum number of orders in the industry, respectively, where N0 ≥ N. min Finally, the final price comparison result S of the j-th repair shop is obtained. ju ; S4.3: Within the preset evaluation period, based on the final price comparison result S of the j-th repair shop. j u The system iterates through price comparison results from various repair shops offering car maintenance and repair services, and then outputs the results via a visualization function r. j The standardized price comparison results are presented in chart form, sorted from largest to smallest, to the car user's client. r j For the output price comparison result of the j-th repair shop, min(S) j u ) and max(S j u S represents the price comparison results from various repair shops. j u The minimum and maximum values, R(S) j u S represents the price comparison result. j u The preset upper limit value; This embodiment specifically explains the use of chart formats such as bar charts and radar charts. Bar charts clearly display the final ranking of different repair shops, while radar charts intuitively reflect the performance of each repair shop in different dimensions such as price, time cost, and quality. Detailed data reports are also provided, showing the specific data and calculation process for each repair shop's various indicators, facilitating in-depth understanding for car users.

[0020] S5: Record the repair shop selections made by car users based on the final price comparison results, perform deviation analysis on the selected repair shop results, and optimize the car repair and maintenance price comparison evaluation process based on the deviation analysis results, including the following steps: S5.1: Within the preset evaluation period, record the repair shop finally selected by the user, compare the ranking results of each repair shop obtained by the car repair and maintenance price comparison evaluation result presentation module, filter out U repair shops whose price comparison results are not the largest, and calculate the selection deviation coefficient η(de), η(de)=U / TU, where TU is the total number of car user maintenance project requests; S5.2: Compare the calculated selection deviation coefficient η(de) with the threshold. If it is less than the threshold, it indicates that the car repair and maintenance price comparison assessment is good. Otherwise, optimize the car repair and maintenance price comparison assessment process. For example, based on the repair shop selected by the car user, analyze the three assessment models of the repair shop and dynamically adjust the weights of the three assessment models.

[0021] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other. In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for comparative evaluation of automobile repair and maintenance prices based on multi-dimensional data fusion, characterized in that: include: S1: Collect price data, time cost data, and repair quality data for car repair and maintenance through the data acquisition interface; S2: Analyze the collected price data, time cost data, and repair quality data respectively, and construct a price comparison evaluation model, a time cost comparison evaluation model, and a repair quality comparison evaluation model; The price comparison evaluation model P1 is P1=exp-[η(p)+η(pp)+η(moy)], where η(p) is the market average price fluctuation coefficient for the same maintenance item, η(pp) is the parts price fluctuation coefficient, and η(moy) is the labor cost difference coefficient. The price comparison time cost assessment model P2 is given by P2=η(t)+η(nd)+η(tg), where η(t) is the response time index, η(nd) is the construction completion rate, and η(tg) is the service efficiency coefficient. The price comparison repair quality assessment model P3 is given by P3=η(nh)+η(CT)+η(DP), where η(nh) is the technical certification coverage rate, η(CT) is the parts compatibility rate, and η(DP) is the maintenance process specification coefficient. S3: Perform two-stage adaptive weight adjustment on the price comparison evaluation model, the time cost comparison evaluation model, and the maintenance quality comparison evaluation model to obtain the optimized adaptive weight adjustment results for each model; The two-stage adaptive weight adjustment includes a first-stage adaptive weight adjustment and a second-stage adaptive weight adjustment. The first stage of adaptive weight adjustment: Based on historical data of each model value, dynamically adjust the base weight W of the I-th model. I ; The second stage adaptive weight adjustment: based on the preference strength Q of the I-th model obtained from the analysis of historical behavior of car users. I The I-th model weight W is obtained through deep optimization. I use ; S4: Based on the adaptive weight adjustment results, the multi-dimensional price comparison comprehensive score is calculated by integrating the price comparison evaluation model, the price comparison time cost evaluation model, and the price comparison repair quality evaluation model. The multi-dimensional price comparison comprehensive score is then corrected to generate a price comparison result with confidence interval, and the final price comparison result is presented to the car user's client. S5: Record the repair shop selection results of the car user client based on the final price comparison results, perform deviation analysis on the selected repair shop results, and optimize the car repair and maintenance price comparison evaluation process based on the deviation analysis results.

2. The method for comparative evaluation of automobile repair and maintenance prices based on multi-dimensional data fusion according to claim 1, characterized in that: In S2, a price comparison and evaluation model is constructed: Using a data platform for the automotive repair and maintenance industry, the price p collected from the j-th repair shop for the same maintenance project is analyzed n1 times, based on the i-th price collected from the j-th repair shop. j,i To obtain the market average price fluctuation coefficient η(p) for the same maintenance and repair project, ,μ(p i () represents the average market price for this maintenance and repair item at the time of the i-th data collection. Then, the price pp for the j-th repair shop in the n2th iteration is collected for the i-th iteration of the price collection for the parts required to be replaced in the maintenance project. j,i The average market price μ(pp) corresponding to the required replacement parts i ), to obtain the parts price fluctuation coefficient η(pp); then collect the j-th repair shop's ith labor cost moy for the maintenance project n3 times. j,i Compared to industry benchmark hourly rates j,0 The labor cost difference coefficient η(moy).

3. The method for comparative evaluation of automobile repair and maintenance prices based on multi-dimensional data fusion according to claim 1, characterized in that: S2 constructs a price comparison time cost assessment model: the time t from the user initiating a maintenance project request to the maintenance merchant's first response after n4 data collections for the j-th maintenance merchant. i After taking the logarithm and standardizing, the response time index η(t) is obtained. Simultaneously, the number of orders (nd) where the repair shop completed the promised work within the promised timeframe for the n4 maintenance projects is calculated to obtain the work completion rate η(nd), where η(nd) = nd / n4; The one-way driving time (tg) from the car user to the repair shop for the n4 maintenance projects is also calculated. i The service efficiency coefficient η(tg) is obtained. .

4. The method for comparative evaluation of automobile repair and maintenance prices based on multi-dimensional data fusion according to claim 1, characterized in that: In S2, a price comparison repair quality assessment model is constructed: The ratio of the number of industry certifications (nh) held by the technicians of the j-th repair shop to the total number of technicians (T_nh) is collected to obtain the technical certification coverage rate η(nh); the set of source types CT for the replacement parts required for the j-th repair shop in the maintenance project is collected n2 times, CT=[ct1,ct2,...,ctm], where m is the number of source types of replacement parts, and ctm is the source of the m-th part; then, the set of part quantities n_CT for the m-th part sources is calculated, n_CT=[n_ct1,n_ct2,...,n_ctm], where n_ctm is the number of parts from the m-th part source, to obtain the part compatibility rate η(CT). n_ctJ represents the quantity of original parts, and n_ctp represents the quantity of branded parts. Through the digital management system of the repair shop, the execution record of the i-th process step in the n5 complete maintenance process is collected and compared with the standard maintenance process specification to obtain the maintenance process specification coefficient η(DP).

5. The method for comparative evaluation of automobile repair and maintenance prices based on multi-dimensional data fusion according to claim 1, characterized in that: In S3, the basic weights W of the I-th model are dynamically adjusted. I , W I，0 P represents the base weights of the i-th model, α is the weight adjustment coefficient, and its value is [0,1]. I (t), P I ,avg、P I ,max and P I ,min represents the current value, historical average value, historical maximum value, and historical minimum value of the I-th model, respectively. I=1 is the price comparison evaluation model, I=2 is the price comparison time cost evaluation model, and I=3 is the price comparison maintenance quality evaluation model.

6. The method for comparative evaluation of automobile repair and maintenance prices based on multi-dimensional data fusion according to claim 1, characterized in that: The I-th model weight W in S3 is deeply optimized. I use Within a preset evaluation period, based on maintenance requests initiated by car users, the proportion of orders in the car user's historical orders that selected the lowest price option is retrieved as the preference strength Q1 of the price comparison evaluation model. Simultaneously, the proportion of orders in which car users selected original parts is retrieved as the preference strength Q3 of the repair quality comparison model. Furthermore, using natural language processing technology, the proportion of orders in which car users inquired about time through the client interaction platform is obtained as the preference strength Q2 of the time cost comparison model. This yields the deeply optimized I-th model weight W. I use , Q I Let β be the preference strength of the i-th model. I Let be the preference coefficient of the i-th model. W I，max This is the preset upper limit for the weights of the I-th model.

7. The method for comparative evaluation of automobile repair and maintenance prices based on multi-dimensional data fusion according to claim 1, characterized in that: The multi-dimensional price comparison comprehensive score in S4 is calculated by integrating the price comparison evaluation model, the price comparison time cost evaluation model, and the price comparison repair quality evaluation model based on the adaptive weight adjustment results. j , P I For the i-th evaluation model value, W I use This refers to the I-th model weight that has undergone deep optimization.

8. The method for comparative evaluation of automobile repair and maintenance prices based on multi-dimensional data fusion according to claim 1, characterized in that: In step S4, a price comparison result with confidence intervals is generated: within a preset evaluation period, the multi-dimensional price comparison comprehensive score is corrected based on the number of service orders for the j-th repair merchant, and a price comparison result S with confidence intervals for the j-th repair merchant is generated. j u , N j 、N0、N min and N max These represent the number of service orders for the j-th repair shop, the industry benchmark number of service orders, the minimum number of orders in the industry, and the maximum number of orders in the industry, respectively; finally, the final price comparison result S for the j-th repair shop is obtained. j u ; S4.3: Within the preset evaluation period, based on the final price comparison result S of the j-th repair shop. j u The system iterates through price comparison results from various repair shops offering car maintenance and repair services, and then outputs the results via a visualization function r. j The standardized price comparison results are presented in chart form, sorted from largest to smallest, to the car user's client. r j For the output price comparison result of the j-th repair shop, min(S) j u ) and max(S j u S represents the price comparison results from various repair shops. j u The minimum and maximum values, R(S) j u S represents the price comparison result. j u The preset upper limit value.

9. The method for comparative evaluation of automobile repair and maintenance prices based on multi-dimensional data fusion according to claim 1, characterized in that: The S5 implementation includes the following steps: S5.1: Within the preset evaluation period, record the repair shop finally selected by the user, compare the ranking results of each repair shop obtained by the car repair and maintenance price comparison evaluation result presentation module, filter out U repair shops whose price comparison results are not the largest, and calculate the selection deviation coefficient η(de), η(de)=U / TU, where TU is the total number of car user maintenance project requests; S5.2: Compare the calculated selection deviation coefficient η(de) with the threshold. If it is less than the threshold, it indicates that the car repair and maintenance price comparison assessment is good; otherwise, the car repair and maintenance price comparison assessment process is optimized.

Images