Artificial intelligence recommendation method, system and storage medium based on auto repair ERP

Abstract

The present invention discloses an artificial intelligence recommendation method, system and storage medium based on auto repair ERP, which relates to the field of data management technology, including an auto repair ERP management platform receiving a vehicle maintenance request, determining the maintenance type, and marking the second-class maintenance and the first-class maintenance respectively; outputting a recommendation list according to the maintenance type, and sorting the recommendation list, obtaining the credit value of the auto repair shop and the reliability of the repair service evaluation of the auto repair shop; receiving a selected instruction, issuing an order containing maintenance information, receiving order confirmation information, and forwarding the auto repair shop information if the auto repair shop confirms the order within a preset time, otherwise outputting the recommendation list again until the maintenance request is completed or accepted by the auto repair shop. The present invention aims to solve the problems that it is difficult for car drivers to obtain high-quality auto repair services, that auto repair data cannot be shared and utilized, that the degree of informatization and digitization is low, that the accessibility of high-quality auto repair services is poor, and that maintenance technology and experience cannot be summarized and modeled.

Description

[0001] This application is a divisional application. The application number of the original application is 202111299154.1, the application date is November 4, 2021, and the name of the invention is: Artificial intelligence automobile maintenance recommendation method, system and storage medium based on ERP. Technical Field

[0002] The present invention relates to the technical field of data management, and in particular to an artificial intelligence recommendation method, system and storage medium based on auto repair ERP. Background Art

[0003] The domestic automobile market economy is growing rapidly. Almost every household has a car. At the same time, services such as automobile maintenance and repair have also emerged. The automobile repair industry has become a new development focus. With the growth of automobile consumption in my country, the number of cars and the automobile aftermarket are growing. The difficulty of maintaining a car is also a problem faced by many car owners. Traditional auto repair shops are extensively managed and lack effective management tools. It is difficult to meet customers' demand for high-quality repair services. In addition, there is a lack of reference standards for the quality and price of automobile repair services. Information asymmetry makes it difficult for drivers and repair shops to establish general trust, which also increases transaction costs. It is difficult for auto repair shops to maintain a stable customer base, which is not conducive to the development of the automotive aftermarket industry. On the other hand, a large amount of automobile maintenance data cannot be accumulated and shared, and the data cannot be used to improve services. In the Internet era, the data resources of the auto repair industry have not been effectively utilized, and information technology and artificial intelligence technology have not yet been combined with auto repair services to improve service quality and user experience. The auto repair industry is in urgent need of the application of new technologies. It is necessary to help drivers establish connections with high-quality auto repair shops to obtain better services, and to use automobile maintenance big data to improve the maintenance quality and efficiency of auto repair shops, and to use data resources to leverage the advantages of information technology to solve problems, especially in the area of ​​paint repair, which is highly perceived by Chinese car owners. Traditional maintenance methods rely on the personal skills of maintenance technicians, and the demand for technological innovation is even greater.

[0004] The Chinese invention patent application with publication number CN110781381A discloses a data verification method based on a neural network, which receives automobile maintenance data to be verified sent by a user, extracts automobile damage pictures and repair plans in the automobile maintenance data, matches the repair plan with the standard repair plan in the repair plan database, and determines whether there is a standard repair plan consistent with the repair plan in the repair plan database. This method can only be used to verify the maintenance plan or data given by the auto repair shop itself. The maintenance data verification model cannot be updated based on existing maintenance data and new data, nor can it recommend better maintenance plans based on the state of the car. The driver still cannot know the maintenance progress, quality and service differences of the auto repair shop, which makes it difficult to meet the technical innovation needs of the auto repair shop. Summary of the invention

[0005] In view of the technical problems in existing automobile maintenance and paint repair, such as it is difficult for drivers to obtain high-quality auto repair services, the inability to share and utilize automobile repair data of auto repair shops, the low degree of informatization and digitization, the poor accessibility of high-quality auto repair services, and the failure to summarize and model the maintenance technology and experience contained in big data, the artificial intelligence recommendation method, system and storage medium based on auto repair ERP provided by the present invention can effectively establish connections between drivers and auto repair shops, efficiently allocate maintenance requests, recommend excellent maintenance service providers to drivers, and use maintenance data to establish a paint repair recommendation model, and can use the auto repair ERP management platform to realize cloud-based storage and sharing of data.

[0006] To solve the above problems, the present invention provides a method for recommending automobile maintenance, comprising the following steps:

[0007] The auto repair ERP management platform receives a vehicle repair request and determines the repair type. If the repair request includes paint damage, the repair request is marked as a second-class repair; otherwise, the repair request is marked as a first-class repair.

[0008] Outputting a recommended list of auto repair shops and / or technicians according to the maintenance type, wherein the recommended list is sorted according to the auto repair shop work order records and the distance to the auto repair shop, or the recommended list is sorted according to the recommendation degree of the auto repair shop;

[0009] Receive the selected instruction, issue an order containing maintenance information, receive order confirmation information, and if the auto repair shop confirms the order within a preset time, forward the auto repair shop information, otherwise output the recommended list again until the maintenance request is completed or accepted by the auto repair shop.

[0010] Optionally, the method further includes collecting image information of the vehicle, analyzing the image information to determine the paint color of the vehicle, and if the paint color of the vehicle is different from the color configured when the vehicle leaves the factory, it is recommended to repaint using a colorimetric method.

[0011] Optionally, if the paint color of the vehicle is the color configured when the vehicle leaves the factory, vehicle information is collected, and the historical maintenance records of the vehicle are obtained from the auto repair ERP management platform and the third-party platform. When the vehicle maintenance type is Category II maintenance, if there is a paint maintenance record for the vehicle containing a paint formula, a maintenance plan is recommended based on the paint formula. Otherwise, the paint formula is obtained from the paint maintenance records of other vehicles of the same model as the vehicle as the basic recommended maintenance plan. If there is no paint maintenance record for the vehicle and other vehicles of the same model, it is recommended to repaint using the colorimetric method.

[0012] Optionally, it also includes collecting the vehicle's maintenance record and progress, collecting the data recorded by the colorimeter during the vehicle's maintenance, and if the color difference is less than a preset value, forwarding the vehicle's maintenance progress and photos of the vehicle after maintenance; otherwise, it prompts that the color difference does not meet the standard, receives a maintenance service evaluation or authorization instruction, and sends the vehicle's maintenance record to a third-party platform according to the authorization instruction.

[0013] Optionally, it also includes obtaining the credit value of the auto repair shop and the credibility of the repair service evaluation of the auto repair shop. If the credit value and the credibility are both greater than the mean or within a preset range, the value obtained by multiplying the color difference compliance rate of the auto repair shop by the credit value and the value obtained by multiplying the repair service praise rate by the credibility are added to output the auto repair shop recommendation degree.

[0014] Optionally, it also includes collecting image information of the maintenance station where the vehicle is located in the auto repair shop, analyzing the image information to determine whether the vehicle has left the maintenance station while in the auto repair shop; if so, determining whether it is an abnormal operation based on the time and duration of the vehicle leaving the maintenance station, and determining whether there is any abnormal movement in the image information of the vehicle while at the maintenance station; if not, determining whether there is any abnormal movement in the image information of the vehicle while at the maintenance station.

[0015] Optionally, if the paint color of the vehicle is the color configured when the vehicle leaves the factory, vehicle information is collected, and historical maintenance records of the vehicle are obtained from the auto repair ERP management platform and the third-party platform. When the vehicle maintenance type is Category II maintenance, if there is no paint maintenance record containing the paint formula for the vehicle, but there are paint maintenance records of other vehicles of the same model as the vehicle, then the paint maintenance records of the other vehicles are obtained to construct a training data set, wherein the training data set includes a historical feature matrix and a paint maintenance plan of the other vehicles, and the historical feature matrix is ​​associated with parameter information of the vehicle. The training data set is input into a paint maintenance recommendation model for training, and the paint maintenance recommendation model is a multi-hidden layer neural network model.

[0016] Optionally, if the paint color of the vehicle is the color configured when the vehicle leaves the factory, vehicle information is collected, and the historical maintenance records of the vehicle are obtained from the auto repair ERP management platform and the third-party platform. When the vehicle maintenance type is Category II maintenance, if there is no paint maintenance record containing the paint formula for the vehicle, but there are paint maintenance records of other vehicles of the same model as the vehicle, then the maintenance service evaluation and return repair records of the other vehicles are obtained to construct a verification data set, or the cumulative lighting time and watering time of the other vehicles are obtained to construct a verification data set, and the verification data set is obtained by data cleaning based on the credit value of the auto repair shop and / or the credibility of the maintenance service evaluation.

[0017] To achieve the purpose of the present invention, the present invention also provides an automobile maintenance recommendation system, including a driver end, an auto repair shop end, an auto repair ERP management platform and a third-party platform, the driver end and the auto repair shop end are both connected to the auto repair ERP management platform, the auto repair ERP management platform is connected to the third-party platform, the auto repair ERP management platform stores vehicle maintenance data and a paint maintenance recommendation model, and the automobile maintenance recommendation system executes the above-mentioned automobile maintenance recommendation method when it is running.

[0018] To achieve the purpose of the present invention, the present invention also provides a computer-readable storage medium, in which a computer program or a code set is stored. When the computer program or the code set is executed by a processor, some or all of the steps of the automobile maintenance recommendation method are implemented.

[0019] Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects:

[0020] (1) Drivers can access the automobile repair recommendation system to obtain timely and efficient recommendations for auto repair shops. The auto repair shop recommendation list may also include technician information and auto repair shop recommendation scores. The auto repair shop recommendation scores are derived from the auto repair shop's technical indicators reflecting service quality and customer evaluation indicators;

[0021] (2) Drivers only need to put forward maintenance requirements, and the matching and allocation of maintenance services will be solved by the automobile maintenance recommendation system through the allocation of resources on the automobile maintenance ERP management platform, which reduces the burden on drivers to select and evaluate maintenance service providers;

[0022] (3) The data of the auto repair shop can be stored and shared by the auto repair ERP management platform. The auto repair shop can obtain technical and management support and feedback from the auto repair ERP management platform. The auto repair shop only needs to update the work order record and upload the maintenance data. It has more time and motivation to provide good maintenance services, and can obtain reasonable evaluation and intelligent recommendations in the automobile maintenance recommendation system, making it easier to obtain a stable customer base and good reviews.

[0023] (4) Using Internet communication technology to connect drivers' maintenance needs and auto repair shops' maintenance services, and storing big data resources on the cloud-based auto repair ERP management platform is more suitable for the application of cloud computing technology and neural network algorithms, achieving lightweight deployment on both the driver side and the auto repair shop side, which is in line with the trend of technological development and is more convenient for application and implementation in traditional industries such as auto repair;

[0024] (5) Since the credit value of the auto repair shop and the credibility of the repair service evaluation are dynamically updated values, and some of the data comes from outside the automobile repair recommendation system, the training of the paint repair recommendation model can also achieve continuous reinforcement learning. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic diagram of the structure of the automobile maintenance recommendation system proposed in an embodiment of the present invention;

[0026] Figure 2 This is an information interaction flow chart of the automobile maintenance recommendation method proposed in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0028] In the description of the present invention, it should be noted that, unless otherwise clearly stipulated and limited, the terms "installation", "connection" and "connection" should be understood in a broad sense. For example, it can be a fixed connection, an integral connection, or a detachable connection; it can be a mechanical connection or an electrical connection, or it can be the internal connection of two elements; it can be a direct connection, or it can be indirectly connected through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to the specific circumstances.

[0029] like Figure 1 The automobile maintenance recommendation system shown includes a driver end, an automobile repair shop end, an automobile maintenance ERP management platform and a third-party platform. The driver end and the automobile repair shop end are connected to the automobile maintenance ERP management platform, and the automobile maintenance ERP management platform is connected to the third-party platform. The automobile maintenance ERP management platform stores vehicle maintenance data and a car paint maintenance recommendation model. When the automobile maintenance recommendation system is running, some or all steps of the automobile maintenance recommendation method proposed in the present invention are executed. The automobile maintenance ERP management platform includes a resource server for storing vehicle maintenance data, a vehicle model database, a car paint database and a formula database, and a formula query module, a formula data acquisition module and a formula data comparison module connected to the resource server. The formula query module is used to receive query conditions and obtain existing formulas in the formula database and output the initial formula; the formula data acquisition module is used to collect the initial formula and the final formula; the formula data comparison module is used for comparing and analyzing the initial formula and the final formula, and comparing and analyzing the final formula and the existing formula. The vehicle model database contains information such as vehicle model, vehicle body color, year, paint supplier and vehicle picture.

[0030] The formula query module is connected to the formula matching analysis module, the formula matching analysis module is connected to the formula database, and the formula data acquisition module is connected to the terminal device, which includes a touch screen, a keyboard, an automatic color mixing machine, a barcode scanner or an online electronic scale. Each formula data generated by the paint mixing operation in the formula database contains information such as the work order number, vehicle model, vehicle body color, vehicle factory time, formula generation time, repair part, formula generation location or repair location, paint mixing operator name or code, paint mixing operator points, formula cost, verified mark, comment information, number of times adopted and paint brand.

[0031] The recipe query module includes an input module, a retrieval module and a sorting module. The input module is used to receive query conditions, and the query conditions include vehicle model information, vehicle body color and target color information of maintenance parts.

[0032] The auto repair ERP management platform also includes a login authentication module. Both the driver side and the auto repair shop side are connected to the login authentication module. The login authentication module includes a login receiving module, a password verification module and an identity authentication module. The login receiving module is used to receive user information and user passwords. User information includes user names and user groups. User groups can be classified according to the user's formula management permissions. User groups include system administrators, system users and paint operators; the password verification module contains an encryption algorithm that can encrypt user passwords and user names. The identity authentication module is connected to a biometric information recognition module. The biometric information recognition module is used to process biometric information such as fingerprint information, facial information, iris information or vein information.

[0033] On the premise of setting up work orders and work order numbers on the auto repair ERP management platform, after the paint operator logs into the auto repair recommendation system, he retrieves relevant formula data in the formula database according to the target color information, vehicle model information and body color information through the retrieval module of the formula query module. The sorting module is connected to the paint database located on the resource server. The sorting module includes a default sorting unit and a re-sorting unit. The default sorting unit is used to sort and display the formula query results according to preset rules. The preset rules include but are not limited to giving priority to displaying formulas with larger paint stocks or displaying formulas with smaller paint stocks later after associating the stock information of the paint database with the formula data. The preset rules are used to recommend formulas with sufficient paint to system users; the re-sorting unit is used to sort and display the formula query results according to the sorting type specified by the user. The sorting types specified by the user include time sorting, relevance sorting, paint operator sorting, weather sorting, cost sorting, number of times adopted sorting and distance sorting. Time sorting is based on the current month of formula query and adjacent months. The closer the formula generation time is to the current date of formula query, the more reference value it has. If the formula data is not from the current month or adjacent months, the formula data generated in the same month and adjacent months of previous years will be displayed first. This is because the effect of the repair paint operation is affected by the temperature and air humidity. Formula query and repair paint operation are usually required to be carried out on the same day, and the operating conditions of the formula data of the current month are relatively close. Distance sorting is based on the location of the formula query as the starting point to calculate the distance to the location where the formula data in the query result is generated. The closer the distance, the similar the regional climate and weather, the closer the operating conditions, and the more reference value. Relevance sorting matches the formula data information and sorts it by the number of matches of vehicle model, body color, car factory time, formula generation time, repair part and paint brand. The formula data with a higher matching number will be displayed first.

[0034] After receiving the query conditions, the recipe query module transmits the query conditions to the recipe matching analysis module. The recipe matching analysis module determines whether the query conditions match the recipe data stored in the recipe database. If yes, it outputs the recipe query result. If not, it prompts that there is no relevant recipe. The recipe data in the recipe query result is decrypted and output by the decryption module. Whether to decrypt is determined based on the user's recipe management authority. The recipe data provided for matching analysis by the recipe matching analysis module does not include the recipe data of the problematic recipe library.

[0035] The auto repair ERP management platform also includes a communication module for connecting the driver side and the auto repair shop side. The communication module is connected to the Internet. The auto repair ERP management platform can obtain weather information and timestamps through the Internet. The weather information and timestamps can be associated with work orders to ensure the accuracy of the information.

[0036] The auto repair ERP management platform also includes a formula cost analysis module connected to the formula query module. The formula cost analysis module is connected to the car paint database located on the resource server. The car paint database contains the brand, price and inventory information of masterbatch, varnish, primer and spray auxiliary materials. When the car paint inventory in the car paint database is insufficient, the auto repair ERP management platform can determine whether to remind the auto repair shop to replenish the car paint based on the car paint consumption or sales data in the region combined with the user's common car paint types, and can also adjust the car paint supply price in the ERP system based on the car paint consumption or sales data. The formula cost analysis module is used to calculate the cost of the formula data displayed in the formula query results, especially the material cost, so that the formula query results can be sorted by cost, which is convenient for the paint mixing operator to evaluate the availability of the formula from a cost perspective.

[0037] like Figure 2 The automobile maintenance recommendation method shown includes the following steps: the auto repair ERP management platform receives a vehicle maintenance request, and at the same time or after receiving the vehicle maintenance request, requests and obtains the vehicle location information, and encrypts the vehicle location information; determines the maintenance type, if the maintenance request includes paint damage, then marks the maintenance request as a second-category maintenance, otherwise marks the maintenance request as a first-category maintenance; outputs a recommended list of auto repair shops and / or technicians according to the maintenance type, the recommended list is sorted according to the auto repair shop work order record and the distance to the auto repair shop, or the recommended list is sorted according to the recommendation degree of the auto repair shop; receives a selected instruction, issues an order containing maintenance information, receives order confirmation information, if the auto repair shop confirms the order within a preset time, then forwards the auto repair shop information, otherwise outputs the recommended list again until the maintenance request is completed or accepted by the auto repair shop.

[0038] In other embodiments, it also includes collecting image information of the vehicle and determining whether the vehicle has paint damage through an image analysis module. When submitting a vehicle repair request, the driver can mark the paint damage and / or upload the image information of the vehicle. If the driver marks that a traffic accident has occurred, it will prompt to upload the image information of the vehicle and provide an expedited option or directly mark the repair request as an expedited order. If the auto repair ERP management platform issues an expedited order containing repair information, the preset time allowed for the auto repair shop to confirm the order will be shortened by a certain percentage.

[0039] In other embodiments, it also includes collecting image information of the vehicle, analyzing the image information to determine the paint color of the vehicle, and if the paint color of the vehicle is different from the color configured when the vehicle leaves the factory, it is recommended to repaint using a colorimetric method.

[0040] In other embodiments, if the paint color of the vehicle is the color configured when the vehicle leaves the factory, vehicle information is collected, and the historical maintenance record of the vehicle is obtained from the auto repair ERP management platform and the third-party platform. When the vehicle maintenance type is Class II maintenance, if there is a paint maintenance record of the vehicle containing a paint formula, a maintenance plan is recommended based on the paint formula. Otherwise, the paint formula is obtained from the paint maintenance records of other vehicles of the same model as the vehicle as the basic recommended maintenance plan, and the vehicle age is selected from the paint maintenance records of other vehicles of the same model as the vehicle through vehicle information matching. The closer vehicle is selected, and vehicles repaired in the same auto repair shop or by the same technician are given priority; or, the light sensor record and rain sensor record of the vehicle are collected, and the light sensor record and rain sensor record of other vehicles of the same model as the vehicle are obtained, and the light sensor record and rain sensor record respectively contain the accumulated light exposure time and the water spraying time, and the accumulated light exposure time and the water spraying time of the vehicle are compared with those of the other vehicles, and the vehicle with similar time is selected; if there is no paint repair record for the vehicle and other vehicles of the same model, it is recommended to repaint by colorimetry. When the paint formula is obtained from the paint repair records of other vehicles of the same model as the vehicle as the basic recommended repair plan, the evaluation of the recommended repair plan by the auto repair shop is received.

[0041] In other embodiments, when collecting the light sensor records and rain sensor records of the vehicle, if the light sensor records and rain sensor records cannot be obtained, vehicles located in the same city or other adjacent cities in the same latitude are selected from the paint maintenance records of other vehicles of the same model as the vehicle through vehicle information matching, and priority is given to vehicles repaired in the same auto repair shop or by the same technician.

[0042] Colorimetric repainting is to use the existing formula queried as the initial formula for paint mixing and color matching. The repair paint prepared according to the initial formula is sprayed on the sample. After the sample is dry, the color of the sample is compared with the color of the vehicle body. The color difference data can be tested with a colorimeter. If necessary, the initial formula can be fine-tuned until the final formula that meets the color difference standard is obtained. If the final formula is different from the existing formula stored in the formula database, it is determined whether the final formula meets the paint mixing rules. If yes, the final formula is subject to a formula compliance review. If not, the final formula is included in the low-distribution formula library and the paint mixing operator points are adjusted. If the final formula is consistent with other known existing formulas, the adopted record of the existing formula is updated. Paint mixing rules include the intermediate color principle, that is, any two non-complementary colors mixed to obtain an intermediate color; the complementary color principle, that is, color A and the complementary color of color A mixed to obtain black; the substitution principle: that is, masterbatch A can be replaced by masterbatch B, and the color mixing effects of masterbatch A and masterbatch B are similar. The brightness of the masterbatch can be increased by adding white or reduced by adding black. Paint mixing rules also include paint mixing requirements determined according to the characteristics of the masterbatch.

[0043] Determine whether the final formula has passed the formula compliance review. If yes, the final formula will be included in the formula library to be verified; if not, the final formula will be included in the problem formula library, and the paint mixing operator's score will be adjusted. Paint mixing operators include manual and machine operators. Machine paint mixing operators include automatic color mixing machines or color mixing systems. When the paint mixing operation is performed manually, the formula compliance review is particularly necessary. The formula compliance review includes but is not limited to comparing the operation data collected by the terminal device connected to the formula data collection module with the final formula data to determine whether there are contradictions or inconsistencies, or judging whether it is obviously unreasonable based on the paint inventory information in the paint database, such as the inventory of a certain ingredient in the final formula is obviously insufficient.

[0044] The automobile maintenance recommendation method also includes collecting the vehicle's maintenance record and progress, collecting the data recorded by the colorimeter during the vehicle's maintenance, and if the color difference is less than a preset value, forwarding the vehicle's maintenance progress and photos of the vehicle after maintenance, otherwise prompting that the color difference does not meet the standard, receiving maintenance service evaluation or authorization instructions, and sending the vehicle's maintenance record to a third-party platform according to the authorization instructions.

[0045] The automobile repair recommendation method also includes obtaining the credit value of the auto repair shop and the credibility of the repair service evaluation of the auto repair shop. If the credit value or credibility is less than a preset value, the auto repair shop is not included in the recommendation list; if the credit value and credibility are both greater than the average or within a preset range, the auto repair shop recommendation is output by adding the value obtained by multiplying the color difference compliance rate of the auto repair shop by the credit value and the value obtained by multiplying the repair service praise rate by the credibility. The auto repair shop recommendation is mapped to the auto repair shop attribute, and the auto repair shop recommendation is provided when outputting the auto repair shop and / or technician recommendation list. The auto repair shop recommendation and the auto repair shop name can be displayed adjacent to each other at the same time, and the car paint brand, type, certification information and corresponding service warranty period used by the auto repair shop can also be displayed for users to refer to when selecting an auto repair shop. Since the auto repair shop recommendation depends on the auto repair shop credit value and the credibility of the repair service evaluation to a certain extent, it helps to encourage the auto repair shop to maintain and improve its own credit value, avoid the adverse effects of false information, and also improve the user experience.

[0046] The automobile maintenance recommendation method also includes collecting image information of the maintenance station where the vehicle is located in the auto repair shop, analyzing the image information to determine whether the vehicle leaves the maintenance station during the auto repair shop, and if so, judging whether it is an abnormal operation according to the time and duration of the vehicle leaving the maintenance station, and judging whether the image information of the vehicle during the maintenance station has abnormal movement; if not, judging whether the image information of the vehicle during the maintenance station has abnormal movement. If it is judged that there is an abnormal movement, the image within the duration of the abnormal movement and the image within a period of time before and after the abnormal movement are automatically recorded, preferably the image within 30 to 120 seconds before and after the abnormal movement occurs. The abnormal movement includes abnormal maintenance behavior, such as personnel staying near or inside the vehicle for a long time, people smoking or using fire near the vehicle, the vehicle being hit by foreign objects, and the maintenance station where the vehicle is located cannot correspond to the maintenance status. The number of abnormal operations and the number of abnormal movements are both used as calculation parameters for the credit value of the auto repair shop. The calculation parameters also include the number of times the auto repair shop confirms the order and then cancels the order.

[0047] In other embodiments, if the paint color of the vehicle is the color configured when the vehicle leaves the factory, the vehicle information is collected, and the historical maintenance record of the vehicle is obtained from the auto repair ERP management platform and the third-party platform. When the vehicle maintenance type is the second-class maintenance, if there is no paint maintenance record containing the paint formula of the vehicle, but there is a paint maintenance record of other vehicles of the same model as the vehicle, then the paint maintenance records of other vehicles are obtained to construct a training data set, and the training data set includes the historical feature matrix and the paint maintenance plan of other vehicles. The historical feature matrix is ​​associated with the parameter information of the vehicle, and the training data set is input into the paint maintenance recommendation model for training. The paint maintenance recommendation model is a multi-hidden layer neural network model. The multi-hidden layer neural network model is a pre-trained model, and the pre-trained model is obtained by training the initial model by inputting simulated training data and / or training data provided by a third-party platform. The training method of the paint maintenance recommendation model is to select one of the hidden layers as the starting point for layer-by-layer training until the last layer, and then select one layer from the untrained hidden layers as the starting point for layer-by-layer training until the last layer, until at least one round of traversal is completed; it helps to avoid error transmission and solidification of training effects.

[0048] When the data volume of the training data set does not reach the set value, the training data set also includes simulated training data. When the data volume of the training data set does not reach the set value, the simulated training data is first selected to input the paint repair recommendation model for training. This helps to reduce error transmission and accelerate convergence.

[0049] In other embodiments, if the paint color of the vehicle is the color configured when the vehicle leaves the factory, vehicle information is collected, and the historical maintenance records of the vehicle are obtained from the auto repair ERP management platform and the third-party platform. When the vehicle maintenance type is Class II maintenance, if there is no paint maintenance record containing the paint formula for the vehicle, but there are paint maintenance records of other vehicles of the same model as the vehicle, then the maintenance service evaluation and return repair records of the other vehicles are obtained to construct a verification data set, or the cumulative lighting time and watering time of the other vehicles are obtained to construct a verification data set, and the verification data set is obtained by data cleaning based on the credit value of the auto repair shop and / or the credibility of the maintenance service evaluation.

[0050] In other embodiments, the numerical calculation of the recommendation degree of an auto repair shop also includes the number of times the auto repair shop has been shared. The driver-side operation can recommend the auto repair ERP management platform or the driver side or the auto repair shop to other drivers or new users. The number of times the auto repair shop has been shared is multiplied by the conversion coefficient, and the value obtained by multiplying the color difference compliance rate of the auto repair shop by the credit value and the value obtained by multiplying the repair service praise rate by the credibility. The conversion coefficient can be 0.5, and it is limited that after the number of times the auto repair shop has been shared exceeds 20 times, it will no longer be calculated by multiplying the number of times the auto repair shop has been shared by the conversion coefficient, but will be replaced by a fixed value, such as 10.

[0051] In other embodiments, when outputting a list of recommended auto repair shops and / or mechanics, they may be sorted and displayed by the recommendation level of the auto repair shop. The recommendation level of the auto repair shop may be presented as a calculated numerical value, or may be graded as high, medium, and elementary, or may be presented as nine levels from 1 to 9. The graded presentation makes it easier for users to view auto repair shops of different levels, providing a better user experience.

[0052] In other embodiments, while outputting a list of recommended auto repair shops and / or mechanics based on the repair type, the historical repair information of the vehicle is matched based on the vehicle information and an auto repair shop that has serviced the vehicle is recommended.

[0053] In other embodiments, when the vehicle maintenance type is Class II maintenance, while outputting a recommended list of auto repair shops and / or mechanics based on the maintenance type, auto repair shops within a preset radius of the vehicle's location are also screened based on the auto repair shops' color difference compliance rate and the praise rate of maintenance service evaluations.

[0054] In other embodiments, vehicle information is collected, and historical maintenance records of the vehicle are obtained from the auto repair ERP management platform and the third-party platform. When the vehicle maintenance type is the second-class maintenance, if there is no paint maintenance record containing the paint formula of the vehicle, but there is a paint maintenance record of other vehicles of the same model as the vehicle, and the data volume of the paint maintenance records of the other vehicles exceeds the set value, then the paint maintenance records of the other vehicles are obtained to construct a practical training set, the maintenance service evaluation and return repair records of the other vehicles are obtained to construct a customer training set, and the color difference data of the other vehicles that meet the color difference standards are obtained to construct a verification set, and the verification set is obtained by data cleaning based on the credit value of the auto repair shop. The practical training set and the customer training set are successively input into the multi-hidden layer neural network model for training to obtain the target paint maintenance recommendation model, and then the verification set is used to verify the target paint maintenance recommendation model.

[0055] In other embodiments, the practical training set and the customer training set are respectively input into a multi-hidden layer neural network model for training to obtain two target vehicle paint repair recommendation models, which are then verified using the verification set, and the two target vehicle paint repair recommendation models are compared to obtain better model parameters and vehicle paint repair recommendation models.

[0056] The embodiment of the present invention further provides a computer storage medium, in which a computer program or a code set is stored, and when the computer program or the code set is executed by a processor or a computer, the computer readable storage medium is controlled to execute some or all steps of the automobile maintenance recommendation method proposed by the present invention. The computer includes a personal computer, a tablet computer, a laptop computer, a PDA, a mobile phone or a server.

[0057] The computer-readable storage medium may be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared or semiconductor system, device or device, a magnetic card, an IC card, a USB flash drive, an SD card, etc. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any combination of the above. The computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, device or device. The computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any combination of the above. A computer-readable signal medium may also be any computer-readable storage medium other than a computer-readable storage medium that can send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, optical cable, radio frequency signals, etc.

[0058] Attached Figure 2 The information interaction flow chart in shows the possible architecture, function and operation of the system, method and computer program of the embodiment of the present invention. Each square frame of the flow chart can represent a part of a module, program segment or code, and the part of the above-mentioned module, program segment or code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative schemes, the functions marked in the square frame can also occur in an order different from that marked in the accompanying drawings. For example, the program modules or executable instructions represented by two consecutively represented square frames can actually be executed substantially in parallel, and they can also be executed in reverse order according to functional requirements sometimes.

[0059] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments may still be modified, or some or all of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for recommending automobile maintenance, characterized in that: The following steps are involved: The auto repair ERP management platform receives a vehicle repair request and determines the repair type. If the repair request includes paint damage, the repair request is marked as a second-class repair; otherwise, the repair request is marked as a first-class repair. Outputting a recommended list of auto repair shops and / or technicians according to the maintenance type, wherein the recommended list is sorted according to the auto repair shop work order records and the distance to the auto repair shop, or the recommended list is sorted according to the recommendation degree of the auto repair shop; Obtaining the credit value of the auto repair shop and the credibility of the repair service evaluation of the auto repair shop. If the credit value or the credibility is less than a preset value, the auto repair shop is not included in the recommendation list; Receive the selected instruction, issue an order containing maintenance information, receive order confirmation information, and if the auto repair shop confirms the order within a preset time, forward the auto repair shop information, otherwise output the recommended list again until the maintenance request is completed or accepted by the auto repair shop; Collect image information of the vehicle, analyze the image information to determine the paint color of the vehicle, if the paint color of the vehicle is different from the color configured when the vehicle leaves the factory, it is recommended to repaint by colorimetry; if the paint color of the vehicle is the color configured when the vehicle leaves the factory, collect vehicle information, obtain the historical maintenance record of the vehicle from the auto repair ERP management platform and the third-party platform, when the vehicle maintenance type is Class II maintenance, if there is no paint maintenance record containing the paint formula of the vehicle, but there is a paint maintenance record of other vehicles of the same model as the vehicle, then obtain the paint maintenance record of the other vehicles to construct a training data set, the training data set includes the historical feature matrix and paint maintenance plan of the other vehicles, the historical feature matrix is ​​associated with the parameter information of the vehicle, and the training data set is input into the paint maintenance recommendation model for training, and the paint maintenance recommendation model is a multi-hidden layer neural network model.

2. The automobile maintenance recommendation method according to claim 1, if the paint color of the vehicle is the color configured when the vehicle leaves the factory, vehicle information is collected, and the historical maintenance record of the vehicle is obtained from the auto repair ERP management platform and the third-party platform. When the vehicle maintenance type is Class II maintenance, if there is a paint maintenance record of the vehicle containing a paint formula, a maintenance plan is recommended based on the paint formula. Otherwise, the paint formula is obtained from the paint maintenance records of other vehicles of the same model as the vehicle as the basic recommended maintenance plan. If there is no paint maintenance record for the vehicle and other vehicles of the same model, it is recommended to repaint using the colorimetric method.

3. The automobile maintenance recommendation method according to claim 1 or 2 further includes collecting the vehicle's maintenance record and progress, collecting the data recorded by the colorimeter during the vehicle's maintenance, and if the color difference is less than a preset value, forwarding the vehicle's maintenance progress and photos of the vehicle after maintenance, otherwise prompting that the color difference does not meet the standard, receiving a maintenance service evaluation or an authorization instruction, and sending the vehicle's maintenance record to a third-party platform according to the authorization instruction.

4. The automobile repair recommendation method according to claim 3, if the credit value and the credibility are both greater than the mean or within a preset range, the recommendation degree of the auto repair shop is output by adding a value obtained by multiplying the color difference compliance rate of the auto repair shop by the credit value and a value obtained by multiplying the repair service praise rate by the credibility.

5. The automobile maintenance recommendation method according to claim 1 or 2 further includes collecting image information of the maintenance station where the vehicle is located in the auto repair shop, analyzing the image information to determine whether the vehicle has left the maintenance station while in the auto repair shop, and if so, determining whether it is an abnormal operation based on the time and duration of the vehicle leaving the maintenance station, and determining whether the image information of the vehicle while at the maintenance station has abnormal movement; if not, determining whether the image information of the vehicle while at the maintenance station has abnormal movement.

6. The automobile maintenance recommendation method according to claim 1, if the paint color of the vehicle is the color configured when the vehicle leaves the factory, vehicle information is collected, and the historical maintenance record of the vehicle is obtained from the auto repair ERP management platform and the third-party platform. When the vehicle maintenance type is Class II maintenance, if there is no paint maintenance record containing the paint formula for the vehicle, but there is a paint maintenance record of other vehicles of the same model as the vehicle, then the maintenance service evaluation and return repair record of the other vehicles are obtained to construct a verification data set, or the accumulated illumination time and watering time of the other vehicles are obtained to construct a verification data set, and the verification data set is obtained by data cleaning based on the credit value of the auto repair shop and / or the credibility of the maintenance service evaluation.

7. Automobile maintenance recommendation system, characterized in that: It includes a driver side, an auto repair shop side, an auto repair ERP management platform and a third-party platform, the driver side and the auto repair shop side are both connected to the auto repair ERP management platform, the auto repair ERP management platform is connected to the third-party platform, the auto repair ERP management platform includes a communication module for connecting the driver side and the auto repair shop side, the communication module is connected to the Internet, the auto repair ERP management platform stores vehicle maintenance data and a paint maintenance recommendation model, and the auto repair recommendation system executes the auto repair recommendation method described in any one of claims 1-6 when running.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program or a code set, and when the computer program or the code set is executed by a processor, the automobile maintenance recommendation method according to any one of claims 1 to 6 is implemented.

Images