Vehicle maintenance matching management method and system based on mobile maintenance device

Abstract

The application belongs to the technical field of automobile service, and provides a vehicle maintenance matching management method and system based on a mobile maintenance device. The method comprises the following steps: during the provision of maintenance services to the reserved vehicle at the reserved location, the mobile maintenance device sends a first maintenance inquiry signal to the target area; a first response signal is received in response to the first maintenance inquiry signal, and a plurality of first target vehicles are screened based on the first response signal; a second maintenance inquiry signal is sent to each first target vehicle; a plurality of second target vehicles are determined based on a second response signal received in response to the second maintenance inquiry signal, a maintenance agreement for each second target vehicle is automatically generated based on the corresponding second response signal, and an expected maintenance time is determined. The scheme of the application can realize optimal allocation of mobile maintenance service resources, and improve the economic benefits and customer satisfaction of mobile maintenance services.

Description

Technical Field

[0001] This invention relates to the field of automotive service technology, and more specifically, to a vehicle maintenance matching management method and system based on a mobile maintenance device. Background Technology

[0002] With the continuous growth of car ownership, mobile vehicle maintenance services have received widespread attention due to their convenience.

[0003] Existing mobile maintenance devices typically operate on a user appointment system. This means that based on a user's pre-submitted maintenance request and designated location, the device plans its route and arrives at the specified location to provide vehicle maintenance services. However, this traditional service model has significant drawbacks: after completing a single appointment, the mobile maintenance device often needs to return empty or proceed to the next appointment location, remaining idle during this period and failing to fully utilize surrounding potential service resources. Furthermore, because the service scope is limited to pre-booked users, the service recipients are singular and scattered, resulting in high operating costs, low equipment utilization, and poor overall economic efficiency. In addition, the reliance on user-initiated appointments limits service coverage and business expansion capabilities, making it difficult to scale up mobile maintenance services. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes a vehicle maintenance matching management method and system based on a mobile maintenance device, in order to solve at least one of the aforementioned technical problems.

[0005] In a first aspect, the present invention provides a vehicle maintenance matching management method based on a mobile maintenance device, comprising the following steps:

[0006] During the process of providing maintenance services to the scheduled vehicle at the designated location, the mobile maintenance device sends a first maintenance inquiry signal to the target area; the target area is determined based on the number of maintenance services that the mobile maintenance device can provide.

[0007] Receive a first response signal in response to the first maintenance inquiry signal, filter out a number of first target vehicles based on the first response signal, and send a second maintenance inquiry signal to each first target vehicle, including the appointment location and maintenance item information;

[0008] Based on the second response signal received in response to the second maintenance inquiry signal, several second target vehicles are determined, and a maintenance agreement for each second target vehicle is automatically generated based on the corresponding second response signal, and the expected maintenance time is determined.

[0009] In a second aspect, the present invention provides a vehicle maintenance matching management system based on a mobile maintenance device. The system includes a processing device and a storage device. The storage device stores a computer program, and the processing device runs the computer program to perform the following steps:

[0010] During the process of providing maintenance services to the scheduled vehicle at the designated location, the mobile maintenance device sends a first maintenance inquiry signal to the target area; the target area is determined based on the number of maintenance services that the mobile maintenance device can provide.

[0011] Receive a first response signal in response to the first maintenance inquiry signal, filter out a number of first target vehicles based on the first response signal, and send a second maintenance inquiry signal to each first target vehicle, including the appointment location and maintenance item information;

[0012] Based on the second response signal received in response to the second maintenance inquiry signal, several second target vehicles are determined, and a maintenance agreement for each second target vehicle is automatically generated based on the corresponding second response signal, and the expected maintenance time is determined.

[0013] In a third aspect, the present invention provides an electronic device comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the computer program, when executed by the processor, implements the method as described in any of the preceding claims.

[0014] In a fourth aspect, the present invention provides a computer storage medium storing a computer program that can be executed by a processor to implement the method as described in any of the preceding claims.

[0015] In a fifth aspect, the present invention provides a computer program product comprising a computer program executable by a processor to implement the method as described in any of the preceding claims.

[0016] This invention first filters response signals based on maintenance capabilities and vehicle location, identifies the target vehicle, and sends detailed inquiry signals to accurately filter out invalid requests. Then, based on feedback, it determines the final service recipient, automatically generates an agreement, and intelligently plans the maintenance schedule. This invention enables optimal allocation of mobile maintenance service resources, improving the economic efficiency and customer satisfaction of mobile maintenance services. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic flowchart of a vehicle maintenance matching management method based on a mobile maintenance device disclosed in an embodiment of the present invention;

[0019] Figure 2 This is a schematic diagram of the joint analysis model disclosed in the embodiments of the present invention;

[0020] Figure 3 This is a schematic diagram of the mobile maintenance device disclosed in the embodiments of the present invention when it is in a truck-carrying state;

[0021] Figure 4 This is a schematic diagram of the mobile maintenance device disclosed in the embodiments of the present invention, which is in a state where it can provide maintenance services as soon as it is placed on the ground and opened.

[0022] Figure 5 This is a schematic diagram of a vehicle maintenance matching management system based on a mobile maintenance device disclosed in an embodiment of the present invention. Detailed Implementation

[0023] The following specific embodiments illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] Furthermore, the technical features involved in the different embodiments of this application described below can be combined with each other as long as they do not conflict with each other.

[0025] like Figure 1 As shown in the figure, an embodiment of the present invention discloses a vehicle maintenance matching management method based on a mobile maintenance device, including the following method steps:

[0026] S10, during the period of providing maintenance services to the reserved vehicle at the reserved location, the mobile maintenance device sends a first maintenance inquiry signal to the target area; the target area is determined based on the number of maintenance services that the mobile maintenance device can provide.

[0027] As the mobile maintenance device arrives at the scheduled location and provides maintenance services to the reserved vehicle, it simultaneously sends a first maintenance inquiry signal to the target area. This first maintenance inquiry signal is used to inquire whether nearby vehicles require maintenance services. In this way, the mobile maintenance device can proactively explore potential vehicle maintenance needs in the surrounding area, breaking through the traditional limitation of only serving reserved vehicles.

[0028] Meanwhile, mobile maintenance units are limited by factors such as the type of maintenance equipment, the quantity of consumables carried, and the duration of operation, resulting in a significant upper limit to their actual maintenance capabilities. For example, equipped only with basic oil change equipment and consumables and lacking sophisticated engine testing instruments, they cannot handle complex engine fault diagnosis and repair projects. Furthermore, the number of vehicles that can be served within a limited working time is also limited by the number of technicians and the time required for each service session. If mobile maintenance units indiscriminately send inquiry signals to all surrounding vehicles, they are prone to receiving a large number of response signals exceeding their capacity. This not only leads to some vehicles' needs not being met in a timely manner but also causes service delays, resource waste, and seriously affects user experience and service reputation.

[0029] To address this, the present invention configures the mobile maintenance device to send a first maintenance inquiry signal only to vehicles within the aforementioned target area, and the size of the target area is determined based on the number of maintenance services the mobile maintenance device can provide. For example, if the mobile maintenance device can only provide simple tire changes and oil changes, the target area is set to a smaller range (e.g., 500m) to focus on nearby vehicles; if the mobile maintenance device is equipped with more comprehensive equipment and can handle a variety of routine maintenance items, the target area is appropriately expanded (e.g., 1000m). It is understood that the target area can be gradually expanded if there is a lack of sufficient responses.

[0030] This precise area delineation method based on maintenance capabilities can effectively expand the service scope, ensure that the equipment responds to needs efficiently within acceptable limits, achieve optimal allocation of service resources, and improve overall service quality and customer satisfaction.

[0031] S20: Receive a first response signal in response to the first maintenance inquiry signal, filter out a number of first target vehicles based on the first response signal, and send a second maintenance inquiry signal to each first target vehicle, including the appointment location and maintenance item information.

[0032] After receiving the first response signal in response to the first maintenance inquiry signal, the mobile maintenance device filters out several target vehicles. Understandably, the filtering process requires consideration of factors such as the distance of the vehicles, the order of response, and the true intent of the user for the mobile maintenance service; this will be analyzed further later.

[0033] After identifying the primary target vehicles, the mobile maintenance device sends a second maintenance inquiry signal to each of these vehicles. This signal includes the appointment location and maintenance service details, such as the available services and pricing. Vehicle owners can then decide whether to accept the service based on their vehicle's condition, distance from the appointment location, and the services offered by the mobile maintenance device. This ensures precise service matching and reduces communication costs and service scheduling conflicts caused by unclear needs.

[0034] It should be noted that all vehicles sending the first response signal to the mobile maintenance device are equipped with an intelligent in-vehicle system. This system has a wireless signal receiver, which can receive the first and second maintenance inquiry signals sent by the mobile maintenance device. The vehicle owner can respond to these inquiries automatically or manually, thus confirming the mobile maintenance appointment. The mobile maintenance device and the wireless signal receiver can interact via mobile network signals, ZigBee signals, or radio.

[0035] S30: Based on the received second response signal in response to the second maintenance inquiry signal, a number of second target vehicles are determined, and a maintenance agreement for each second target vehicle is automatically generated based on the corresponding second response signal, and the expected maintenance time is determined.

[0036] After sending a second maintenance inquiry signal to each of the first target vehicles, if the first target vehicle does indeed require maintenance (e.g., providing specific maintenance items), it will send a second response signal. At this point, the mobile maintenance device selects several qualified second target vehicles from these vehicles that genuinely intend to receive service. These vehicles are the final recipients of the mobile maintenance device's service. The conditions here include, but are not limited to, the feedback sequence number of the second response signal being less than a preset sequence number (representing the maximum maintenance capacity, for example, 8 vehicles; meaning the first 8 vehicles to provide a second response signal are identified as second target vehicles).

[0037] Subsequently, the mobile maintenance device automatically generates maintenance agreements for each second target vehicle based on the corresponding second response signal. The maintenance agreement includes key information such as service items, service prices, and the rights and obligations of both parties. This information is automatically generated based on the needs expressed by the vehicle owner in the second response signal and the service capabilities of the mobile maintenance device, ensuring that the agreement content accurately matches the needs of both parties, while improving the efficiency of agreement drafting and reducing errors that may occur during manual operation. The maintenance agreement and the expected maintenance time are then sent to the corresponding second target vehicle.

[0038] When determining the expected maintenance time, the mobile maintenance unit comprehensively considers factors such as the service progress of the scheduled vehicle, the location of each secondary target vehicle, and the duration required for the service items, and rationally plans the service sequence and schedule. On the one hand, this allows car owners to have a clear expectation of the service time, making it easier for them to arrange their own schedules. On the other hand, it ensures that the mobile maintenance unit efficiently completes all services within the limited working time, avoiding time conflicts and resource waste during the service process, achieving optimal allocation of service resources, and further improving service efficiency and customer satisfaction.

[0039] This invention first filters response signals based on maintenance capabilities and vehicle location, identifies the target vehicle, and sends detailed inquiry signals to accurately filter out invalid requests. Then, based on feedback, it determines the final service recipient, automatically generates an agreement, and intelligently plans the maintenance schedule. This invention enables optimal allocation of mobile maintenance service resources, improving the economic efficiency and customer satisfaction of mobile maintenance services.

[0040] As an example, the mobile maintenance device sends a first maintenance inquiry signal to the target area, including:

[0041] The vehicle model and maintenance items targeted by the maintenance service are determined, and historical maintenance data corresponding to the vehicle model and maintenance items are obtained from the database. Based on the historical maintenance data, the expected maintenance time is predicted.

[0042] Based on the expected maintenance time, the remaining maintenance time for the current working period is determined. If the remaining maintenance time is higher than the time threshold, the mobile maintenance device sends a first maintenance inquiry signal to the target area; otherwise, the mobile maintenance device does not send a first maintenance inquiry signal to the target area.

[0043] When performing mobile maintenance services, since the duration of each work session is limited, blindly sending inquiry signals to accept too many maintenance services can easily lead to maintenance services exceeding the time limit and a decline in quality. Therefore, it is necessary to reasonably control the timing of sending the first maintenance inquiry signal.

[0044] In this embodiment, the mobile maintenance device first determines the model and maintenance items of the scheduled vehicle for the currently performed maintenance service. Simultaneously, a large number of maintenance records are pre-stored in the database, involving vehicles of different brands and models, and various maintenance items. Using the vehicle model and maintenance items as an index, a search is performed in this database to obtain multiple corresponding historical maintenance records. The actual maintenance time contained in these matched historical maintenance records is processed to obtain an equivalent maintenance time (e.g., the average maintenance time), which is then used as the expected maintenance time.

[0045] The remaining maintenance time is then calculated based on the total duration of the current working period (e.g., 8:00-18:00 daily), and compared with a preset time threshold. If the remaining maintenance time is higher than the threshold (e.g., 2 hours), it indicates that there is sufficient time to undertake additional services, and the mobile maintenance device will send an inquiry signal; otherwise, it will not send an inquiry signal.

[0046] This setup avoids time conflicts and resource waste caused by service overload, allowing mobile maintenance devices to plan services reasonably within a set time, effectively improving resource utilization efficiency and service reliability, and ensuring service quality and customer satisfaction.

[0047] As an example, the selection of several first target vehicles based on the first response signal includes:

[0048] Calculate the response duration of each of the first response signals. If the response duration is higher than the first duration threshold, then the vehicle corresponding to the response signal is identified as the first target vehicle.

[0049] If the response duration is lower than a first duration threshold but higher than a second duration threshold, an automatic response probability analysis is performed on the virtual response signal group using a joint analysis model; wherein each virtual response signal in the virtual response signal group contains motion state data of the corresponding vehicle; if the automatic response probability obtained from the analysis is lower than the probability threshold, the vehicle corresponding to the first response signal is determined as the first target vehicle.

[0050] Because car maintenance services typically take a long time, generally 1-3 hours, the maintenance services that a mobile maintenance device can provide in a single work session are actually limited, for example, 5-8 vehicles. Therefore, this invention prioritizes vehicles with clear maintenance needs in an initial sorting queue, identifying them as the first target vehicles. Simultaneously, since there are differences between manual and automatic responses, automatic responses cannot accurately reflect the maintenance needs of the vehicle occupants. Without proper differentiation and screening, this can easily lead to wasted service resources and reduced service efficiency. Therefore, this invention uses a hierarchical screening mechanism based on response time.

[0051] First, calculate the response time for each response signal. The response time refers to the time difference between the time the first maintenance inquiry signal is sent and the time the first response signal is sent (carried in the first response signal), or the time difference between the time the first maintenance inquiry signal is sent and the time the first response signal is received. The shorter the time difference, the faster the response of the corresponding vehicle.

[0052] When the response time exceeds the first time threshold, the first response signal can be identified as a manual reply, indicating that the person in the vehicle actively pays attention to and manually confirms the service request. Such vehicles are directly identified as the first target vehicles.

[0053] When the response time is less than the second time threshold, i.e. the response is too fast, it can be judged as an automatic reply. At this time, it cannot reflect the real maintenance needs of the people in the vehicle and will not be included in the first target vehicle.

[0054] When the response duration falls between a first duration threshold and a second duration threshold, it becomes difficult to determine whether the first response signal is a manual or automatic response. To address this, the present invention further constructs a joint analysis model. This model combines the motion state data of the corresponding vehicle contained in the virtual response signal group to analyze the probability that the first response signal is an automatic response, i.e., the automatic response probability. If the automatic response probability is lower than the probability threshold, it indicates a higher probability of a manual response, and the corresponding vehicle is identified as the first target vehicle; conversely, it indicates a higher probability of an automatic response, and the corresponding vehicle is not identified as the first target vehicle for the time being. It is understandable that the vehicle's motion state data reflects the complexity of the vehicle's motion. When the motion complexity is high (e.g., high-speed driving in complex traffic flow), it is actually difficult for occupants, especially the driver, to quickly and manually respond to the first maintenance inquiry signal. Conversely, when the motion complexity is low (e.g., the vehicle is parked in a parking lot, or the vehicle is stopped and waiting to start in congested traffic), the probability of occupants quickly and manually responding to the first maintenance inquiry signal is much higher. This invention utilizes the relationship between the vehicle's motion state and the manual quick response to the first maintenance inquiry signal to analyze the probability of automatic response to the first response signal whose response time is lower than a first time threshold but higher than a second time threshold.

[0055] Through the precise identification and screening mechanism described in this embodiment, vehicles that truly require maintenance can be effectively identified, reducing invalid service matching, improving the efficiency of service resource utilization, and ensuring the accuracy and efficiency of mobile maintenance services.

[0056] It should be noted that when the mobile maintenance device detects a first response signal with a response duration lower than the first duration threshold but higher than the second duration threshold, it further continuously sends a special signal to the vehicle that issued the first response signal (the first response signal includes the vehicle's ID). Each time the corresponding vehicle receives the special signal, it will automatically send a virtual response signal to the mobile maintenance device. The virtual response signal only includes the vehicle's motion status data, such as location information, vehicle speed information, orientation information, and the corresponding timestamp, and does not contain any maintenance interaction information or vehicle identity information, thus avoiding the leakage of privacy information.

[0057] As an example, the joint analysis model includes a spatiotemporal graph neural network unit, a multimodal Transformer unit, and a fusion processing unit; then, the automatic response probability analysis of the virtual response signal group using the joint analysis model includes:

[0058] The spatiotemporal graph neural network unit constructs each motion state data into a single-node temporal graph structure, and extracts the first spatiotemporal feature based on the feature vector of each node.

[0059] The multimodal Transformer unit is used for:

[0060] Multimodal and location encoding enhancements are performed on the first spatiotemporal features, response duration sequence, and environmental context to obtain modal feature data; the environmental context includes geographic location encoding and timestamp; a multi-head self-attention mechanism is used to capture long-distance dependencies between different modal feature data to generate a semantic representation of the vehicle's response intent;

[0061] Furthermore, the first spatiotemporal feature is encoded to obtain the second spatiotemporal feature; historical response data of the same type, geographical region, and time period as the target vehicle are extracted from the database to generate a historical feature vector; and the association weight between the second spatiotemporal feature and the historical feature vector is calculated through an attention mechanism.

[0062] The fusion processing unit integrates the first spatiotemporal features, the semantic representation of the vehicle's response intent, and the associated weights to calculate the probability of automatic response.

[0063] Determining the probability of automatic vehicle response involves multi-source heterogeneous data such as motion state, response duration, and geographical location, with complex temporal relationships and semantic dependencies between the data. Traditional single models have significant limitations in handling such problems: they cannot effectively extract the temporal features of vehicle motion, are difficult to integrate semantic information from multi-dimensional data, and cannot utilize historical behavior patterns to improve judgment accuracy, resulting in low judgment accuracy and affecting the target vehicle selection efficiency of mobile maintenance devices.

[0064] To address the aforementioned problems, this invention proposes a joint analysis model composed of spatiotemporal graph neural network units, multimodal Transformer units, and fusion processing units, such as... Figure 2As shown in the diagram, the spatiotemporal graph neural network unit constructs a single-node temporal graph structure from vehicle motion state data. Through deep analysis of the feature vectors of each node, it extracts the first spatiotemporal feature reflecting the vehicle's motion pattern, effectively capturing the temporal changes in vehicle motion. The multimodal Transformer unit enhances the first spatiotemporal feature, response duration sequence, and environmental context through multimodal and positional encoding. It utilizes a multi-head self-attention mechanism to mine long-distance dependencies between different modalities, generating a semantic representation of the vehicle's response intent. Simultaneously, this unit performs secondary encoding on the first spatiotemporal feature to obtain the second spatiotemporal feature and incorporates historical response data of the same type, geographical region (at least including the target region, and significantly larger than the target region), and time period as the current vehicle. By calculating the association weights between the historical feature vectors and the second spatiotemporal feature, the accuracy of the analysis is further improved. Finally, the fusion processing unit integrates the first spatiotemporal feature, the semantic representation of the vehicle's response intent, and the historical association weights, and calculates the automatic response probability.

[0065] The aforementioned joint analysis model, through the collaborative operation of multiple units, achieves multi-level and multi-dimensional in-depth analysis and fusion of vehicle data, significantly improving the accuracy of automatic response probability judgment, helping mobile maintenance devices to accurately screen target vehicles, optimize service resource allocation, and improve overall service efficiency and quality.

[0066] The specific processing procedure is illustrated below with an example:

[0067] The spatiotemporal graph neural network unit constructs a single-node time-series graph structure based on the target vehicle's own motion state data. The single-node time-series graph uses a single vehicle as its core, with the node feature vectors corresponding to different times as nodes, thus forming a graph structure reflecting the vehicle's motion state changes over time. In the single-node time-series graph structure, the node feature vectors corresponding to each node (each node has its own corresponding node feature vector) integrate the vehicle's key motion information within a continuous time window, namely, position coordinates (reflecting the vehicle's trajectory), rate of change of speed (reflecting acceleration or deceleration trends), and direction angle (indicating changes in driving direction). The feature vectors of all nodes together constitute the first spatiotemporal feature.

[0068] The edge weights between nodes employ a dynamic decay mechanism: ω t.t+Δt =exp(-λ·Δt), where λ is a trainable parameter and Δt is the time difference. This design allows the model to focus more on recent behavior.

[0069] The feature vectors of each node are arranged in chronological order to form a sequence X = [x1, x2, ..., x]. T The spatiotemporal dependencies are extracted using a graph neural network (GNN) to generate the first spatiotemporal feature H. s1 It captures the continuity and regularity of vehicle movement.

[0070] Multimodal Transformer unit:

[0071] (1) Multimodal feature encoding

[0072] The first spatiotemporal feature X, the response duration sequence T, and the environmental context C (geographic location encoding and timestamp) are independently encoded to generate the first feature representation.

[0073] Using the sine and cosine functions PE = (pos, 2i) = sin(pos / 10000) 2i / d H is the first feature representation. m1 By endowing it with time-series awareness, a second feature representation H is obtained. m2 This addresses the issue of the Transformer architecture lacking temporal modeling capabilities. Here, pos represents the position, i.e., the location of the data within the entire sequence at a given moment; i is the index value used to iterate through the dimension of the feature vectors; and d represents the dimension of the feature vectors.

[0074] A multi-head self-attention mechanism is used to compute dependencies between different modalities, generating a semantic representation H of the vehicle's response intent. intention For example, by associating features such as "morning rush hour", "parking lot location" with "low-speed driving", the likelihood of manual operation can be inferred.

[0075] (2) For the first spatiotemporal feature H s1 Perform Transformer encoding to generate a more abstract second spatiotemporal feature H s2 This will further refine the patterns and laws governing vehicle movement.

[0076] Historical response data of the same type, geographical region, and time period as the current vehicle are extracted from the database to construct a historical feature vector.

[0077] The association weights between current features and historical patterns are calculated using an attention mechanism:

[0078]

[0079] If the current behavior is highly similar to the historical manual response pattern (e.g., a car owner frequently responds manually when driving at low speed during the morning rush hour), the association weight increases, reducing the probability of automatic response.

[0080] Fusion processing unit:

[0081] The first spatiotemporal feature H s1 (Preserving original motion details), semantic representation of response intent H intention (Multimodal association information) and historical association weight α iIntegration, calculating the probability of automatic response through nonlinear transformation:

[0082]

[0083] P = sigmoid(W3·H+b)

[0084] Where σ is the activation function, and W1, W2, W3, and b are trainable parameters. When P is below the probability threshold, it is determined to be a manual response, and the vehicle corresponding to the first response signal is identified as the first target vehicle; otherwise, it is determined to be an automatic response, and the first response signal cannot reflect the actual maintenance needs of the occupants of the vehicle, so it is not included in the first target vehicle for the time being.

[0085] It should be noted that if the first response signal sent by the corresponding vehicle includes a marker for the automatic response mode, its automatic response mode can be directly determined. However, due to the diverse designs of vehicles, the first response signal of many vehicles does not contain this marker. Therefore, it is necessary to use the joint analysis model described above in this invention to infer its automatic response probability.

[0086] As an example, the calculation of the association weights between the second spatiotemporal feature and the historical feature vector through the attention mechanism includes:

[0087] The first association weight between the second spatiotemporal feature and the historical feature vector is calculated using an attention mechanism. The number of records and the record density of the historical response data are calculated, and the weight coefficient is determined based on the number of records. The weight coefficient increases as the number of records and the record density increase.

[0088] The first association weight is corrected using the weight coefficient to obtain the second association weight.

[0089] The aforementioned embodiments introduced historical response data as a reference when analyzing the automatic response probability. However, if the historical response data is insufficient to reflect the general response situation in the region, it may lead to reference errors and affect the accuracy of the final automatic response probability. To address this, the present invention designs a dynamic reference mechanism for historical response data, specifically:

[0090] First, an attention mechanism is used to calculate the first association weight between the second spatiotemporal feature and the historical feature vector. The attention mechanism is a technique that allows the model to focus on important information. Based on factors such as similarity and correlation between the second spatiotemporal feature (reflecting information such as the vehicle's current movement pattern) and the historical feature vector (generated from historical response data of the same type, geographical region, and time period extracted from a database), the first association weight is calculated to measure the degree of association between them. For example, if the current vehicle's movement pattern is very similar to the pattern during a previous human response, the attention mechanism will result in a relatively high first association weight.

[0091] The number of historical response records is calculated, and a weighting coefficient is determined based on this number. The weighting coefficient is positively correlated with the number of records and record density (time density); that is, the larger the number of records and the higher the record density, the larger the weighting coefficient. This is because more and denser historical response data in a region indicates a higher acceptance of mobile maintenance services and a better understanding of the intention behind mobile maintenance services as a primary maintenance inquiry signal. In this case, the probability of occupants responding to the primary maintenance inquiry signal out of curiosity is lower, and the historical response data is more valuable. Conversely, the probability of occupants responding to the primary maintenance inquiry signal out of curiosity is higher, and the historical response data is less valuable.

[0092] The first association weight is corrected using predetermined weighting coefficients to obtain the second association weight. The final second association weight will be used in subsequent fusion processing units to participate in the calculation of the automatic response probability, thereby affecting the judgment of whether the vehicle responds automatically or manually.

[0093] As an example, such as Figure 3 , Figure 4 As shown, the mobile maintenance device is transported by a vehicle and includes a carriage. The carriage includes carriage pillars, several carriage doors, several tool supports, and maintenance equipment arranged on the tool supports; wherein, the carriage pillars are telescopic pillars.

[0094] The process of using the mobile maintenance device of the present invention is roughly as follows:

[0095] The truck carries the mobile maintenance equipment to the designated location. The truck's uprights (e.g., four or more) are lowered to the ground for support, detaching the truck bed from the truck's platform. The truck drives away, and the uprights retract to place the truck bed on the ground. All doors (e.g., four) open, and tool holders (e.g., three) extend outside the truck bed, at which point the vehicle maintenance service can begin.

[0096] For example: The car to be serviced is driven onto the lift, the oil collector collects the engine oil, the tires are removed, the tire changer removes the tires, new tires are installed, the tires are balanced, the tires are installed, a four-wheel alignment is performed, the lift is lowered, and the serviced car leaves; the tool racks in three directions carrying the service equipment are retracted into the cargo box, the four doors in four directions are closed, the cargo box pillars are raised to the predetermined height, the truck is reversed to receive the cargo box, and the mobile service equipment is carried back or to the next service location.

[0097] In addition, the mobile maintenance unit's cabin surface is equipped with solar panels. During the day, the solar panels and energy storage are utilized, and at night they can power electrical equipment and serve as a charging station for vehicles, generating additional economic benefits. Understandably, this also includes a battery, details of which will not be elaborated upon.

[0098] It is understood that the vehicle maintenance matching management system 100 of the present invention is also installed in the passenger compartment, and the specific details will not be elaborated further.

[0099] like Figure 5 As shown, this embodiment of the invention also provides a vehicle maintenance matching management system 100 based on a mobile maintenance device. The system includes a processing device 101 and a storage device 102, wherein the storage device 102 stores a computer program; the processing device 101 runs the computer program to perform the following steps:

[0100] During the process of providing maintenance services to the scheduled vehicle at the designated location, the mobile maintenance device sends a first maintenance inquiry signal to the target area; the target area is determined based on the number of maintenance services that the mobile maintenance device can provide.

[0101] Receive a first response signal in response to the first maintenance inquiry signal, filter out a number of first target vehicles based on the first response signal, and send a second maintenance inquiry signal to each first target vehicle, including the appointment location and maintenance item information;

[0102] Based on the second response signal received in response to the second maintenance inquiry signal, several second target vehicles are determined, and a maintenance agreement for each second target vehicle is automatically generated based on the corresponding second response signal, and the expected maintenance time is determined.

[0103] This invention also provides an electronic device comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the computer program, when executed by the processor, implements the method as described in any of the foregoing embodiments.

[0104] This invention also provides a computer storage medium storing a computer program that can be executed by a processor to implement the methods described in any of the foregoing claims.

[0105] This invention also provides a computer program product comprising a computer program that can be executed by a processor to implement the method as described in any of the foregoing embodiments.

[0106] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0107] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A vehicle maintenance matching management method based on a mobile maintenance device, characterized in that: The methods and steps include the following: During the process of providing maintenance services to the scheduled vehicle at the designated location, the mobile maintenance device sends a first maintenance inquiry signal to the target area; the target area is determined based on the number of maintenance services that the mobile maintenance device can provide. Receive a first response signal in response to the first maintenance inquiry signal, filter out a number of first target vehicles based on the first response signal, and send a second maintenance inquiry signal to each first target vehicle, including the appointment location and maintenance item information; Based on the second response signal received in response to the second maintenance inquiry signal, several second target vehicles are determined, and a maintenance agreement for each second target vehicle is automatically generated based on the corresponding second response signal, and the expected maintenance time is determined. The mobile maintenance device sends a first maintenance inquiry signal to the target area, including: The vehicle model and maintenance items targeted by the maintenance service are determined, and historical maintenance data corresponding to the vehicle model and maintenance items are obtained from the database. Based on the historical maintenance data, the expected maintenance time is predicted. Based on the expected maintenance time, the remaining maintenance time for the current working period is determined. If the remaining maintenance time is higher than the time threshold, the mobile maintenance device sends a first maintenance inquiry signal to the target area; otherwise, the mobile maintenance device does not send a first maintenance inquiry signal to the target area. Based on the first response signal, a number of first target vehicles are identified, including: Calculate the response duration of each of the first response signals. If the response duration is higher than the first duration threshold, then the vehicle corresponding to the response signal is identified as the first target vehicle. If the response duration is lower than a first duration threshold but higher than a second duration threshold, then a joint analysis model is used to perform automatic response probability analysis on the virtual response signal group; wherein, each virtual response signal in the virtual response signal group contains motion state data of the corresponding vehicle; if the automatic response probability obtained from the analysis is lower than the probability threshold, then the vehicle corresponding to the first response signal is determined as the first target vehicle. The joint analysis model includes a spatiotemporal graph neural network unit, a multimodal Transformer unit, and a fusion processing unit; the joint analysis model is then used to perform automatic response probability analysis on virtual response signal groups, including: The spatiotemporal graph neural network unit constructs each motion state data into a single-node temporal graph structure, and extracts the first spatiotemporal feature based on the feature vector of each node. The multimodal Transformer unit is used for: Multimodal and location encoding enhancements are performed on the first spatiotemporal features, response duration sequence, and environmental context to obtain modal feature data; the environmental context includes geographic location encoding and timestamp; a multi-head self-attention mechanism is used to capture long-distance dependencies between different modal feature data to generate a semantic representation of the vehicle's response intent; Furthermore, the first spatiotemporal feature is encoded to obtain the second spatiotemporal feature; historical response data of the same type, geographical region, and time period as the target vehicle are extracted from the database to generate a historical feature vector; and the association weight between the second spatiotemporal feature and the historical feature vector is calculated through an attention mechanism. The fusion processing unit integrates the first spatiotemporal features, the semantic representation of the vehicle's response intent, and the associated weights to calculate the probability of automatic response.

2. The vehicle maintenance matching management method based on a mobile maintenance device according to claim 1, characterized in that: The association weights between the second spatiotemporal feature and the historical feature vector are calculated using an attention mechanism, including: The first association weight between the second spatiotemporal feature and the historical feature vector is calculated using an attention mechanism. The number of records and the record density of the historical response data are calculated, and the weight coefficient is determined based on the number of records. The weight coefficient increases as the number of records and the record density increase. The first association weight is corrected using the weight coefficient to obtain the second association weight.

3. The vehicle maintenance matching management method based on a mobile maintenance device according to claim 1, characterized in that: The mobile maintenance device is transported by a vehicle and includes a carriage. The carriage includes carriage pillars, several carriage doors, several tool supports, and maintenance equipment arranged on the tool supports; wherein, the carriage pillars are telescopic pillars.

4. A vehicle maintenance matching management system based on the method of claim 1, the system comprising a processing device and a storage device, wherein the storage device stores a computer program; characterized in that: The processing device runs the computer program.

Images