A vehicle maintenance management system

Abstract

The present application relates to vehicle operation and maintenance management technical field, specifically to a kind of vehicle maintenance and repair operation management system, including acquisition module, loss calculation module, issue module and decision management module;Acquisition multi-source sensor data, based on the data fluctuation amplitude of continuous data in preset time window, filter generation key operation data according to quantile threshold;Component loss value is calculated according to key operation data;Global resource data is obtained and decomposed, combined with vehicle position coordinate to generate surrounding repair resource data;When component loss value exceeds preset physical safety limit value, preliminary repair scheduling instruction is generated based on surrounding repair resource data and stored in local offline cache, when network recovers, upload cloud end to carry out resource conflict processing, and output maintenance scheduling result;The present application reduces invalid calculation, improves loss judgment pertinence, and guarantees scheduling continuity when communication is limited.

Description

Technical Field

[0001] This invention relates to the field of vehicle operation and maintenance management technology, specifically a vehicle maintenance, repair, and operation management system. Background Technology

[0002] With the continuous development of intelligent vehicle operation and vehicle-to-everything (V2X) applications, the demand for vehicle operation management during maintenance and repair is increasing. In particular, real-time assessment of component wear and tear and timely scheduling of maintenance resources during vehicle operation have become important aspects of ensuring safe vehicle operation and improving maintenance efficiency. How to effectively manage vehicle wear and tear and maintenance scheduling under conditions of continuous vehicle operation, unstable communication environment, and dynamic changes in the distribution of maintenance resources is also a current technical challenge.

[0003] Traditional vehicle maintenance, repair, and operation management methods currently mainly rely on the following approaches: periodic maintenance based on thresholds for the entire vehicle or components, fault early warning based on single sensor data, and centralized cloud-based scheduling relying on real-time network connections.

[0004] However, methods such as regular maintenance, single-sensor early warning, and centralized cloud scheduling all have certain drawbacks. For example, regular maintenance is difficult to reflect the dynamic wear and tear of vehicles during actual operation, which can easily lead to delayed or excessive maintenance. Single-sensor data early warning has limited ability to identify abnormal changes in vehicle components, and when a large amount of raw data is directly involved in processing, it can lead to the consumption of computing resources exceeding the preset threshold and the effective abnormal feature extraction rate being lower than the standard conditions. Centralized cloud scheduling, which relies on real-time network connections, has the problem of not being able to generate maintenance scheduling results in a timely manner when communication is limited. At the same time, when multiple vehicles request maintenance resources at the same time, it is also easy to encounter problems such as untimely handling of resource conflicts and insufficient matching of candidate maintenance sites. Summary of the Invention

[0005] The purpose of this invention is to provide a vehicle maintenance and repair operation management system that avoids the inability to form effective scheduling results when the network is unstable and the scheduling deadlock caused by multiple vehicles competing for the same maintenance resource at the same time. It also solves the problem that the loss judgment lacks a continuous correction mechanism and is out of touch with the actual state in the prior art, and effectively avoids the direct participation of all resources in matching, solving the problems of indiscriminate distribution of resource information and imprecise matching of maintenance resources.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A vehicle maintenance and repair operation management system includes a data acquisition module, a wear calculation module, a data distribution module, and a decision management module.

[0008] The data acquisition module is used to collect multi-source sensor data during vehicle operation, calculate the data fluctuation amplitude based on continuous data within a preset time window, filter the data based on the data fluctuation amplitude, output key operating data, and transmit it to the loss calculation module.

[0009] The loss calculation module is used to receive key operating data, calculate the component loss value of the vehicle based on the preset loss change calculation model, output the component loss value, and transmit it to the decision management module.

[0010] The distribution module is used to acquire global resource data, decompose and process the global resource data, generate surrounding maintenance resource data based on the acquired vehicle location coordinates, and distribute the surrounding maintenance resource data to the decision management module.

[0011] The decision management module includes a local offline cache unit, a network status detection unit, and a network communication interface. It is used to receive component wear values ​​and cache surrounding maintenance resource data. When the component wear value exceeds the preset physical safety limit, it generates a preliminary maintenance scheduling instruction based on the surrounding maintenance resource data and stores it in the local offline cache unit. When the network status detection unit detects network recovery, it uploads the preliminary maintenance scheduling instruction to the cloud through the network communication interface for resource conflict handling and outputs the maintenance scheduling result.

[0012] As a further aspect of the present invention, the multi-source sensor data includes on-board diagnostic system data and chassis vibration data;

[0013] Global resource data includes maintenance site data, spare parts resource data, and road network status data.

[0014] As a further aspect of the present invention, the acquisition module generates key operational data based on multi-source sensor data, specifically including:

[0015] Set a preset time window consisting of continuous sampling points, and extract the continuous sampling time data of each sensor channel as continuous data within the preset time window;

[0016] Features are extracted from the time series change rate and local dispersion of continuous data to obtain the data fluctuation amplitude;

[0017] Arrange the non-zero elements in the data fluctuation range in descending order, and extract the values ​​corresponding to the preset quantiles as the filtering threshold.

[0018] The data fluctuation range is compared with the filtering threshold. If the data fluctuation range is greater than the filtering threshold, the identifier corresponding to the multi-source sensor data is set to the reserved state. If the data fluctuation range is less than or equal to the filtering threshold, the identifier corresponding to the multi-source sensor data is set to the filtered state, thus forming a data validity identifier.

[0019] After aligning the multi-source sensor data with the data validity identifier by timestamp, multiply them element by element to filter out key operational data.

[0020] As a further aspect of the present invention, the loss calculation module generates component loss values ​​based on key operational data, specifically including:

[0021] Read the historical cumulative loss value as the current cumulative loss value, and obtain the preset prediction time interval;

[0022] The current loss increment index is obtained by nonlinearly compressing the energy intensity of effective elements in key operational data.

[0023] The loss increment is calculated based on the current loss increment index, the preset wear correction coefficient, and the prediction time interval.

[0024] Add the current cumulative loss value to the loss increment to generate the component loss value at the next moment;

[0025] When the network is connected, the actual wear value is received from the cloud. The prediction deviation is obtained based on the actual wear value and the component wear value at the next moment. The wear correction coefficient is updated based on the current wear correction coefficient as a historical coefficient, the preset first weight and the preset second weight.

[0026] As a further aspect of the present invention, the distribution module generates surrounding maintenance resource data based on global resource data, specifically including:

[0027] Construct a resource association table that includes maintenance site dimensions, resource dimensions, and road network dimensions;

[0028] The resource association table is decomposed into multiple data tables by splitting fields according to the dimensions of maintenance site, resource, and road network, and site attribute data, spare parts attribute data, and road condition attribute data including road congestion coefficient are extracted.

[0029] Obtain the continuous position coordinates of the vehicle, and calculate the heading angle of the driving trajectory based on the continuous position coordinates;

[0030] Based on continuous location coordinates, driving trajectory heading angle, and site attribute data, spare parts attribute data, and road condition attribute data, generate surrounding maintenance resource data within a preset distance and preset angle range, with the vehicle's current location as the starting point and the driving trajectory heading angle as the center line.

[0031] As a further aspect of the present invention, the decision management module generates preliminary maintenance scheduling instructions based on component wear values ​​and surrounding maintenance resource data, specifically including:

[0032] The component wear value is compared with the preset physical safety limit value. If the component wear value is less than or equal to the preset physical safety limit value, the current operating state of the vehicle is maintained. When the component wear value is greater than the preset physical safety limit value, the calculation of the comprehensive scheduling index for the candidate station is triggered.

[0033] Traverse the candidate stations in the surrounding maintenance resource data, calculate the spatial distance parameter from the vehicle's current location to the candidate station, calculate the estimated arrival time parameter based on the estimated arrival time of the vehicle to the candidate station and the road congestion coefficient obtained from the road condition attribute data, and determine the spare parts matching degree based on the spare parts inventory quantity of the candidate station and the spare parts required by the vehicle.

[0034] Based on spatial distance parameters, estimated arrival time parameters, and spare parts matching degree, a comprehensive scheduling index is calculated, and the optimal maintenance site with the smallest comprehensive scheduling index is selected.

[0035] Generate preliminary maintenance scheduling instructions that include vehicle identification, key operational data, component wear values, and the optimal maintenance site, and store them in the local offline cache unit.

[0036] As a further aspect of the present invention, the decision management module generates a multi-vehicle scheduling strategy based on the preliminary maintenance scheduling instruction, specifically including:

[0037] When the network status detection unit detects that the network has been restored, it uploads the initial maintenance scheduling instructions to the cloud;

[0038] Obtain the latest real resource status of the optimal maintenance site and determine whether the resources of the optimal maintenance site are occupied;

[0039] If the resources of the optimal maintenance station are not occupied, the maintenance resources of the optimal maintenance station will be locked directly for the vehicle.

[0040] If the resources of the optimal maintenance site are occupied, a scheduling priority list is constructed based on the maintenance requests uploaded by multiple vehicles received from the cloud.

[0041] Extract the network outage driving time of the vehicle since the initial maintenance scheduling instruction was generated, call the preset first proportional coefficient and second proportional coefficient, and calculate the queuing priority score based on the component wear value, network outage driving time and the preset first proportional coefficient and second proportional coefficient.

[0042] If the queuing priority score is the highest, then the vehicle will be locked with the best maintenance station for repair resources.

[0043] If the queuing priority score is not the highest, the alternative repair station will be recalculated based on the global resource data stored in the cloud, and a redirection instruction will be output.

[0044] As a further aspect of the present invention, the execution result of the multi-vehicle scheduling strategy includes resource lock confirmation information and redirection instructions;

[0045] The decision management module generates scheduling status identifiers, which include network outage cache markers, network synchronization markers, and resource occupancy markers.

[0046] When the initial maintenance dispatch instruction is stored in the local offline cache unit, a network outage cache marker is generated;

[0047] The network communication interface is used to receive and parse redirection commands and status update information sent from the cloud.

[0048] As a further embodiment of the present invention, the vehicle maintenance and repair operation management system also includes an on-board display terminal, which is used to display a pre-decision guidance interface and a scheduling status interface.

[0049] The pre-decision guidance interface is displayed in a list format, showing the component wear value, the optimal maintenance site, and the estimated arrival time.

[0050] The scheduling status interface displays information in log format, including network outage duration, resource conflict resolution results, and resource lock status.

[0051] The beneficial effects of this invention are:

[0052] 1. This invention collects data from multiple sensors, combines the absolute value of the first difference of continuous data with local variance to calculate the data fluctuation amplitude, and uses quantile thresholds to adaptively extract key operating data. This effectively removes a large amount of steady-state redundant information, reduces the computational burden of data processing, and highlights the abnormal characteristics of components. At the same time, the system continuously extrapolates based on historical cumulative losses and current loss increments, and dynamically updates the wear correction coefficient using real loss values ​​in the cloud when the network is connected. This overcomes the problems of traditional periodic maintenance being divorced from actual operating conditions and the limited early warning capabilities of single sensors, and significantly improves the continuity, pertinence, and accuracy of vehicle component wear status assessment.

[0053] 2. This invention effectively eliminates interfering stations unrelated to the driving direction by splitting global resource data into multi-dimensional fields and combining the vehicle's current location with the heading angle of its driving trajectory to generate directional surrounding maintenance resource data. When wear exceeds the limit, the system comprehensively considers spatial distance parameters, the estimated arrival time parameter based on road congestion coefficient, and spare parts matching degree locally to select the optimal maintenance station with the lowest comprehensive scheduling index and stores the preliminary maintenance scheduling instruction in the local offline cache. This mechanism ensures that the vehicle can still form a high-precision local predictive scheduling result in a timely manner when communication is limited or the environment is unstable, avoiding the problems of scheduling interruption and imprecise matching of candidate resources.

[0054] 3. When the network status detection unit detects network recovery, this invention uploads the locally offline cached scheduling instructions to the cloud for real resource status verification. In the face of resource occupation conflict scenarios, the system innovatively combines the component wear value reflecting the urgency of maintenance with the offline driving time reflecting the offline waiting time to calculate the queuing priority score. Based on this score, the optimal resource is locked, and a redirection instruction is automatically issued to the alternative repair station for vehicles that are not selected. This transparent priority evaluation and suboptimal fallback mechanism effectively solves the deadlock and untimely conflict handling problems that are prone to occur when multiple vehicles request maintenance resources at the same time, ensuring the orderliness and rationality of the overall maintenance scheduling. Attached Figure Description

[0055] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0056] Figure 1 This is a module architecture diagram of the vehicle maintenance and repair operation management system in this application embodiment. Detailed Implementation

[0057] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] Please see Figure 1 A vehicle maintenance and repair operation management system includes a data acquisition module, a wear calculation module, a data distribution module, and a decision management module.

[0059] The data acquisition module is used to collect multi-source sensor data during vehicle operation, calculate the data fluctuation amplitude based on continuous data within a preset time window, filter the data based on the data fluctuation amplitude, output key operating data, and transmit it to the loss calculation module.

[0060] The loss calculation module is used to receive key operating data, calculate the component loss value of the vehicle based on the preset loss change calculation model, output the component loss value, and transmit it to the decision management module.

[0061] The distribution module is used to acquire global resource data, decompose and process the global resource data, generate surrounding maintenance resource data based on the acquired vehicle location coordinates, and distribute the surrounding maintenance resource data to the decision management module.

[0062] The decision management module includes a local offline cache unit, a network status detection unit, and a network communication interface. It is used to receive component wear values ​​and cache surrounding maintenance resource data. When the component wear value exceeds the preset physical safety limit, it generates a preliminary maintenance scheduling instruction based on the surrounding maintenance resource data and stores it in the local offline cache unit. When the network status detection unit detects network recovery, it uploads the preliminary maintenance scheduling instruction to the cloud through the network communication interface for resource conflict handling and outputs the maintenance scheduling result.

[0063] Multi-source sensor data includes on-board diagnostic system data and chassis vibration data; global resource data includes maintenance site data, spare parts resource data, and road network status data.

[0064] The acquisition module generates key operational data based on multi-source sensor data, specifically including: setting a preset time window consisting of continuous sampling points, and extracting continuous sampling time data of each sensor channel as continuous data within the preset time window;

[0065] Features are extracted based on the time series change rate and local dispersion of continuous data to obtain the data fluctuation amplitude; non-zero elements in the data fluctuation amplitude are sorted in descending order, and the values ​​corresponding to the preset quantiles are extracted as the filtering threshold.

[0066] The data fluctuation range is compared with the filtering threshold. If the data fluctuation range is greater than the filtering threshold, the identifier corresponding to the multi-source sensor data is set to the reserved state. If the data fluctuation range is less than or equal to the filtering threshold, the identifier corresponding to the multi-source sensor data is set to the filtered state, thus forming a data validity identifier.

[0067] After aligning the multi-source sensor data with the data validity identifier by timestamp, multiply them element by element to filter out key operational data.

[0068] In practice, the vehicle maintenance and repair operation management system can be deployed on the vehicle, in the roadside communication environment, and in the cloud resource management environment.

[0069] During vehicle operation, the acquisition module synchronously collects data from multiple sensor sources. Among them, the on-board diagnostic system data is used to reflect the operating parameters of key vehicle components, such as speed, temperature, and fault codes, while the chassis vibration data is used to reflect information such as road impact, suspension disturbance, and abnormal vibration of components.

[0070] Since the raw collected data usually has problems such as dense sampling, significant non-stationary fluctuation characteristics, and a high proportion of steady-state information, if all the data is directly sent to the subsequent processing stage, it is easy to increase the computational burden and cause local abnormal features to be masked by a large amount of steady data. Therefore, this embodiment first calculates the fluctuation amplitude of continuous data and then completes the filtering according to the threshold.

[0071] The preset time window refers to the statistical interval consisting of continuous sampling points. The value can be a number of continuous sampling periods, and the specific length can be preset according to the sensor sampling frequency and the vehicle operation rhythm. Continuous data refers to the sampling values ​​of each channel arranged in chronological order within this window.

[0072] The calculation of data fluctuation amplitude is used to identify data that changes significantly in a short period of time and has the significance of indicating component abnormality; firstly, the absolute value of the first difference of continuous data is calculated, and then multiplied by the local variance of the corresponding sensor channel within the window to obtain the data fluctuation amplitude;

[0073] The absolute value of the first-order difference is used to reflect the intensity of change between single samples, and the local variance is used to reflect the degree of dispersion of the channel within the window. Its discrete mathematical expression is:

[0074]

[0075] in, This represents the total number of consecutive sampling points within a preset time window. For a single sample value, The average value of the samples within the window; multiplying the two can enhance the characteristics of channels that change rapidly and fluctuate greatly, thereby extracting abnormal signals that are originally easily covered by noise.

[0076] To avoid the inadequacy of fixed thresholds for adapting to different working conditions, this embodiment uses a quantile method to construct the filtering threshold; the non-zero elements in the data fluctuation range are arranged in descending order, and the value corresponding to the preset quantile is taken as the filtering threshold. The preset quantile can be preset according to the system configuration to control the range of the number of features to be retained.

[0077] In practical applications, the value range of the preset quantile is usually set to 70% to 90%. The specific calibration basis is: extract a large amount of historical sample data generated by the same model of vehicle in a standard operating test environment, and statistically analyze the proportion of high-fluctuation data that can clearly indicate abnormal deterioration of components in the overall data.

[0078] If the test shows that this type of abnormal data generally accounts for only 15% of the total data, then the preset quantile is set to 85%. This quantitative calibration method based on data distribution characteristics avoids the problem of inaccurate filtering caused by blindly taking values ​​based solely on qualitative experience. If the data fluctuation of a certain channel is greater than the filtering threshold, then its corresponding identifier is set to the reserved state.

[0079] If the data is less than or equal to the filtering threshold, it is set to the filtering state, thus forming a data validity identifier. The multi-source sensor data and the data validity identifier are aligned by timestamp and multiplied element by element to obtain the key operating data. This key operating data retains the channel information that is more sensitive to component wear and removes a large amount of steady-state or low-meaning data.

[0080] This embodiment captures the correlation between short-term abnormal changes and component deterioration during vehicle operation by cross-channel fluctuation recognition and adaptive mask filtering, transforming the disadvantage of large data volume but sparse effective information into the advantage of retaining key data according to importance.

[0081] This embodiment can extract key operating data based on the fluctuation amplitude screening mechanism, avoid redundant data from directly entering the subsequent loss calculation stage, reduce invalid calculations and improve the targeting of subsequent component loss values;

[0082] Based on the above embodiments, after receiving key operating data, the loss calculation module calculates the loss of vehicle components; it is used to convert the currently collected abnormal operating characteristics into component loss values ​​in order to determine whether the vehicle has reached the condition that requires maintenance.

[0083] Read the historical cumulative loss value as the current cumulative loss value and obtain the preset prediction time interval; calculate the sum of squares of the non-zero elements in the key operating data and take the logarithm to obtain the current loss increment index.

[0084] Based on the current loss increment index, the preset wear correction coefficient, and the prediction time interval, the loss increment is calculated; the cumulative loss value at the current moment is added to the loss increment to generate the component loss value at the next moment.

[0085] When the network is connected, the actual wear value is received from the cloud, the actual wear value is calculated by subtracting the component wear value at the next moment to predict the deviation, and the wear correction coefficient is updated based on the current wear correction coefficient, the first weight and the second weight.

[0086] Among them, the historical cumulative loss value refers to the loss record accumulated by the vehicle in the previous operating cycle, which is used to support the existing wear foundation; the prediction time interval refers to the time span used to calculate the extension of the current loss to the next cycle, and the value can be preset according to the system operating cycle.

[0087] The current loss increment index is used to characterize the contribution of abnormal intensity in key operational data to loss growth;

[0088] Since key operating data includes data with different physical dimensions and numerical magnitudes, such as speed, temperature and chassis vibration, directly calculating the sum of squares of absolute values ​​will cause dimensional confusion and cause the channels with large numerical magnitudes to completely mask the characteristics of small numerical magnitudes, which violates the self-consistency of comprehensive evaluation of multidimensional physical quantities.

[0089] Before performing the sum of squares calculation, the system pre-configures a dimensionless operation based on the reference constant. This involves dividing each non-zero element of the channel by the preset rated physical reference value for that channel, mapping them uniformly to a dimensionless relative fluctuation ratio, and then performing the sum of squares. Under this mechanism, the sum of squares is used to amplify fluctuation characteristics that exceed the preset amplitude threshold, and the logarithm is used to suppress nonlinear overflow caused by out-of-bounds data.

[0090] The wear correction factor is used to balance the sensitivity of wear growth under different vehicles, different components, and different operating conditions, and to correct the differences in the actual impact of the same abnormal input on the wear growth of different components.

[0091] The calculation of loss increment enables the system to map highly fluctuating sensor changes into comparable loss growth. This embodiment combines the current anomaly intensity and the cumulative loss basis through the loss change calculation model to form a continuous result, avoiding absolute judgment based on a single anomaly. This embodiment can generate component loss values ​​based on the superposition relationship between cumulative loss and loss increment, avoiding misjudgment caused by observing only instantaneous data, and ultimately solving the problem of lack of continuous correction mechanism for loss judgment in the prior art.

[0092] Based on the above embodiments, after the distribution module obtains global resource data, it decomposes the global resource data and generates surrounding maintenance resource data based on the vehicle's location coordinates; this is used to reduce the large-scale resource information in the cloud to a set of available maintenance resources related to the current vehicle's driving direction.

[0093] Construct a resource association table that includes maintenance site dimensions, resource dimensions, and road network dimensions; split the fields according to the maintenance site dimension, resource dimension, and road network dimension, and assign a common-source globally unique identifier key to each original association record during the split calculation. Use the common-source globally unique identifier key as a cross-table mapping pointer to decompose the resource association table into multiple data table combinations, and extract and store the site attribute data, spare parts attribute data, and road condition attribute data separately.

[0094] Obtain the continuous position coordinates of the vehicle and calculate the heading angle of the driving trajectory based on the continuous position coordinates; based on the continuous position coordinates, the heading angle of the driving trajectory, and the site attribute data, spare parts attribute data, and road condition attribute data, generate surrounding maintenance resource data within a preset distance and preset angle range with the vehicle's current position as the starting point and the heading angle of the driving trajectory as the center line;

[0095] Among them, the resource association table is used to put maintenance stations, spare parts inventory and road network status into the same data framework, which facilitates subsequent decomposition; field splitting is used to transform large tables that are coupled together into a combination of independently searchable data tables, reducing the caching pressure on the vehicle end;

[0096] Continuous position coordinates are used to reflect the position changes on the vehicle's travel path, the heading angle is used to describe the central trend of the vehicle's direction of movement, and the preset distance and preset angle range are used to limit the effective area of ​​candidate maintenance resources and prevent stations unrelated to the vehicle's direction of travel from entering the decision set.

[0097] This embodiment transforms massive global resource data into surrounding maintenance resource data that matches the vehicle's current location and driving direction by splitting the resource association table and filtering based on direction constraints. This allows subsequent scheduling to process only a local candidate set, reducing interference from irrelevant stations in scheduling decisions.

[0098] This embodiment can generate surrounding maintenance resource data based on resource decomposition and directional filtering, avoiding excessive processing pressure caused by all resources directly participating in matching, and ultimately solving the problems of indiscriminate distribution of resource information and excessively wide candidate range in the prior art.

[0099] Based on the above embodiments, after the decision management module receives the component wear value and caches the surrounding maintenance resource data, when the component wear value exceeds the preset physical safety limit value, it generates a preliminary maintenance scheduling instruction based on the surrounding maintenance resource data and stores it in the local offline cache unit.

[0100] When the network status detection unit detects network recovery, it uploads preliminary maintenance scheduling instructions to the cloud via the network communication interface to handle resource conflicts and outputs maintenance scheduling results.

[0101] This is used to enable vehicles to complete local predictions even when communication is restricted, and to enter the cloud for conflict verification after the network is restored; the preliminary maintenance dispatch instruction includes vehicle identification, key operating data, component wear values ​​and optimal maintenance sites, which are used as inputs for subsequent conflict handling.

[0102] The preset physical safety limit value is based on the factory-set maximum wear thickness of the target component. Temperature coefficient of the current operating environment The preset ratio of the product is determined, and its quantified analytical expression is as follows:

[0103]

[0104] in, For safety redundancy parameters, the value range is limited to [value range missing]. This distinguishes between physical states that can continue to operate and those that require maintenance; the local offline cache unit is used to save scheduling results generated during network outages to prevent instructions from being lost due to communication interruptions;

[0105] The network status detection unit is used to identify whether communication has been restored, and the network communication interface is used to complete instruction uploading and result reception after restoration;

[0106] This embodiment can realize the initial maintenance scheduling of vehicles based on loss judgment, local resource screening and offline caching mechanism, so as to avoid the inability to form effective scheduling results when the network is unstable, and ultimately solve the problem of scheduling interruption caused by communication limitations in the existing technology.

[0107] In the application example, after the vehicle continuously collects data from the on-board diagnostic system and chassis vibration data during operation, the system first generates key operating data and then calculates component wear values ​​based on this data.

[0108] When the wear value of a component reaches a preset condition, the system generates a preliminary maintenance scheduling instruction from the cached surrounding maintenance resource data. After the network is restored, the instruction is uploaded to the cloud, where resource conflict resolution is completed and the maintenance scheduling result is returned. The vehicle display terminal can then display a pre-decision guidance interface and a scheduling status interface to complete the vehicle maintenance and repair operation management process.

[0109] In a preferred embodiment of the present invention, the loss calculation module generates component loss values ​​based on key operating data, specifically including: reading historical cumulative loss values ​​as the current cumulative loss value and obtaining a preset prediction time interval; performing nonlinear compression expression based on the energy intensity of effective elements in the key operating data to obtain the current loss increment index;

[0110] Based on the current loss increment index, the preset wear correction coefficient, and the prediction time interval, the loss increment is calculated; the cumulative loss value at the current moment is added to the loss increment to generate the component loss value at the next moment.

[0111] When the network is connected, the actual wear value is received from the cloud. The prediction deviation is obtained based on the actual wear value and the component wear value at the next moment. The wear correction coefficient is updated based on the current wear correction coefficient as a historical coefficient, the preset first weight and the preset second weight.

[0112] The distribution module generates surrounding maintenance resource data based on global resource data, specifically including: constructing a resource association table containing maintenance site dimensions, resource dimensions, and road network dimensions; decomposing the resource association table into multiple data table combinations by splitting fields according to maintenance site dimensions, resource dimensions, and road network dimensions, and extracting site attribute data, spare parts attribute data, and road condition attribute data including road congestion coefficients;

[0113] Obtain the continuous position coordinates of the vehicle and calculate the heading angle of the driving trajectory based on the continuous position coordinates; based on the continuous position coordinates, the heading angle of the driving trajectory, and the site attribute data, spare parts attribute data, and road condition attribute data, generate surrounding maintenance resource data within a preset distance and preset angle range with the vehicle's current position as the starting point and the heading angle of the driving trajectory as the center line;

[0114] The decision management module generates preliminary maintenance dispatch instructions based on component wear values ​​and surrounding maintenance resource data. Specifically, it compares the component wear value with a preset physical safety limit value. If the component wear value is less than or equal to the preset physical safety limit value, the current operating status of the vehicle is maintained. When the component wear value is greater than the preset physical safety limit value, the calculation of comprehensive dispatch indicators for candidate stations is triggered.

[0115] Traverse the candidate stations in the surrounding maintenance resource data, calculate the spatial distance parameter from the vehicle's current location to the candidate station, calculate the estimated arrival time parameter based on the estimated arrival time of the vehicle to the candidate station and the road congestion coefficient obtained from the road condition attribute data, and determine the spare parts matching degree based on the spare parts inventory quantity of the candidate station and the spare parts required by the vehicle.

[0116] Based on spatial distance parameters, estimated arrival time parameters, and spare parts matching degree, a comprehensive scheduling index is calculated, and the optimal maintenance site with the smallest comprehensive scheduling index is selected. A preliminary maintenance scheduling instruction containing vehicle identification, key operating data, component wear values, and the optimal maintenance site is generated and stored in the local offline cache unit.

[0117] Based on the above embodiments, the loss calculation module uses the method of superimposing historical cumulative loss value and loss increment in the component loss value to form a continuous loss trajectory.

[0118] Reading historical cumulative wear values ​​as the current cumulative wear value allows the system to inherit the wear foundation already formed by the vehicle, avoiding starting from zero for each judgment; the prediction time interval is used to limit the extension span of wear increment, making it easier to map the current operating characteristics to the wear changes in the next cycle.

[0119] The current loss increment index is obtained by taking the logarithm of the sum of squares of non-zero elements in the key operating data. It is used to compress the energy intensity of abnormal inputs and prevent a small number of extreme values ​​from excessively dominating the loss results. The wear correction coefficient is used to adjust the degree of loss amplification under different components and different vehicle conditions. It is a correction parameter for the loss growth rate.

[0120] The update mechanism of the wear correction coefficient depends on the actual wear value returned by the cloud under network connectivity. The difference between the actual wear value and the component wear value at the next moment is used to obtain the prediction deviation, which reflects the gap between the local and cloud actual results. The current wear correction coefficient is used as the historical coefficient, and then the product of the first weight and the second weight is added to form the new wear correction coefficient.

[0121] The first weight is used to maintain the continuity of historical coefficients, and the second weight is used to introduce newly obtained prediction deviation information so that the correction coefficient can gradually approach the actual wear trend.

[0122] Since historical coefficients and prediction deviations belong to different physical dimensions—the former being a mapping adjustment coefficient and the latter a loss absolute value with actual physical units—the system's preset second weight is a parameter factor with error gain conversion attributes, whose dimensions are strictly defined as follows:

[0123]

[0124] By multiplying the prediction bias by the second weight, the dimensions of the result can be reduced to exactly [value missing]. This allows for a precise match with the dimensions of the historical coefficients.

[0125] For example, the predicted wear value of the current component at the next moment is the equivalent wear depth. The actual loss value returned by the cloud is This indicates a local lag, with a prediction bias of [value missing]. If the historical wear correction factor is The first weight is dimensionless. The second weight is configured as follows The updated wear correction factor is:

[0126]

[0127] The increased correction coefficient after adjustment will accelerate the next local loss growth, thus achieving dynamic closed-loop control.

[0128] This embodiment dynamically adjusts the correction coefficient by feeding back the actual loss value, which can reverse the cloud verification result and use it locally, reducing the accumulation of deviations during long-term operation.

[0129] This embodiment can calculate the component loss value based on cumulative loss, increment and feedback correction, avoiding the amplification of deviation caused by long-term fixed loss model, and finally solving the problem of loss calculation being out of touch with the actual state in the prior art.

[0130] Based on the above embodiments, when the distribution module decomposes and processes global resource data, it first constructs a resource association table, and then splits it into multiple data table combinations according to the maintenance station dimension, resource dimension and road network dimension.

[0131] The resource association table is used to unify maintenance stations, spare parts resources, and road network status into the same data framework; the split station attribute data is used to describe the service capabilities of candidate maintenance stations, spare parts attribute data is used to describe inventory status, and road condition attribute data is used to describe the traffic conditions of candidate routes.

[0132] The vehicle's continuous position coordinates are used to reflect the current position change trend, and the driving trajectory heading angle is used to determine the vehicle's direction of travel. The two together can limit the screening area of ​​surrounding maintenance resource data; the preset distance and preset angle range are used to form a candidate set facing the area in front of the vehicle to avoid including stations that are obviously not in line with the driving direction in the scheduling.

[0133] When the component wear value exceeds the physical safety limit, the decision management module triggers the calculation of the comprehensive scheduling index of the candidate sites. For each candidate site, the spatial distance parameter, the estimated arrival time parameter, and the spare parts matching degree are calculated separately, and then a weighted sum is performed to obtain the comprehensive scheduling index.

[0134] Among them, the spatial distance parameter is used to measure the impact of the distance to the station on the scheduling efficiency; the estimated arrival time parameter is obtained by multiplying the estimated arrival time by the road congestion coefficient, and is used to reflect the impact of road traffic conditions on the arrival time.

[0135] The spare parts matching degree is composed of the ratio of the number of spare parts in stock at the candidate site to the number of spare parts required by the vehicle, which is used to characterize whether the site has the resources required for the current maintenance.

[0136] After weighted summation of the three factors, the station with the smallest comprehensive scheduling index is selected as the optimal maintenance station, and a preliminary maintenance scheduling instruction containing vehicle identification, key operating data, component wear value and the optimal maintenance station is generated.

[0137] It should be noted that since the spatial distance parameter and the estimated arrival time parameter are both negative scheduling indicators where smaller values ​​are better, while the spare parts matching degree is a positive resource indicator where larger values ​​are better, in order to make the three parameters logically consistent with the judgment criterion of selecting the optimal maintenance site with the smallest comprehensive scheduling index, when performing weighted summation, the weights assigned to the spatial distance parameter and the estimated arrival time parameter are preset to positive real numbers, while the weights assigned to the spare parts matching degree are negative real numbers.

[0138] It should be noted that, since the spatial distance parameter, estimated arrival time parameter, and spare parts matching degree are physical quantities with different dimensions and numerical ranges exceeding the preset order of magnitude, they were each normalized before the weighted summation of the comprehensive scheduling indicators. The specific normalization conversion formula uses the range transformation method, that is, setting a certain indicator parameter as... Its maximum value in the current set of candidate sites for surrounding maintenance resources is The minimum value is Then its normalized value conversion formula is:

[0139]

[0140] in, To prevent extremely small constants with a denominator of 0, such as 0.0001, the values ​​of all indicators are uniformly mapped to... The interval; this calculation rule can be clearly defined through specific quantification: assuming the normalized value of the spatial distance assessment of a candidate site is... The corresponding weight is ;

[0141] The normalized value of the estimated arrival time assessment is The corresponding weight is The site's spare parts inventory is: The vehicle needs The original value of the spare parts matching degree was calculated. After substituting the historical maximum and minimum values ​​using the range transformation method, its normalized value is Its corresponding weight is set to The overall scheduling objective function value is:

[0142]

[0143] By assigning negative weights to positive indicators, the comprehensive scheduling objective function value of sites with more spare parts is calculated to a lower range, thereby accurately adapting to the optimization logic of minimizing system scheduling parameters and avoiding the principle conflict caused by the direct accumulation of indicators with different monotonicity.

[0144] This embodiment captures the comprehensive compatibility of candidate sites in terms of location, accessibility, and inventory through the collaborative calculation of spatial distance, arrival time, and spare parts matching degree, transforming the matching that relies solely on distance parameters into precise screening based on current maintenance conditions;

[0145] This embodiment can generate preliminary maintenance scheduling instructions based on comprehensive scheduling indicators, avoiding site selection bias caused by relying on a single indicator, and ultimately solving the problem of insufficient matching of maintenance resources in the prior art; in the application embodiment, when the vehicle runs along the driving trajectory and continuously acquires location coordinates, the system will synchronously update the surrounding maintenance resource data.

[0146] When the wear and tear of a component exceeds the physical safety limit, the system immediately calculates the optimal maintenance site locally and caches the initial maintenance dispatch instruction. After the network is restored, the instruction is uploaded to the cloud for resource conflict verification, thereby completing vehicle dispatch and resource allocation.

[0147] In a preferred embodiment of the present invention, the decision management module generates a multi-vehicle scheduling strategy based on the preliminary maintenance scheduling instruction, specifically including: uploading the preliminary maintenance scheduling instruction to the cloud when the network status detection unit detects network recovery; obtaining the latest real resource status of the optimal maintenance site and determining whether the resources of the optimal maintenance site are occupied;

[0148] If the resources of the optimal maintenance site are not occupied, the maintenance resources of the optimal maintenance site are directly locked for the vehicle; if the resources of the optimal maintenance site are occupied, a scheduling priority list is constructed based on the maintenance requests uploaded by multiple vehicles received from the cloud.

[0149] Extract the network outage driving time of the vehicle since the initial maintenance scheduling instruction was generated, call the preset first proportional coefficient and second proportional coefficient, and calculate the queuing priority score based on the component wear value, network outage driving time and the preset first proportional coefficient and second proportional coefficient.

[0150] If the queuing priority score is the highest, the vehicle will be locked with the optimal maintenance station for repairs; if the queuing priority score is not the highest, the alternative maintenance stations will be recalculated based on the global resource data stored in the cloud, and a redirection instruction will be output.

[0151] The execution results of the multi-vehicle scheduling strategy include resource lock confirmation information and redirection instructions; the decision management module generates scheduling status identifiers, which include network outage cache markers, network synchronization markers, and resource occupancy markers;

[0152] When the initial maintenance dispatch instruction is stored in the local offline cache unit, a network outage cache marker is generated; the network communication interface is used to receive and parse the redirection instructions and status update information sent from the cloud.

[0153] The vehicle maintenance and repair operation management system also includes an on-board display terminal, which is used to display the pre-decision guidance interface and the scheduling status interface;

[0154] The pre-decision guidance interface displays information in a list format, including component wear values, optimal maintenance sites, and estimated arrival times; the scheduling status interface displays information in log format, including network outage travel time, resource conflict resolution results, and resource lock status.

[0155] Based on the above embodiments, when the network is restored, the decision management module uploads the preliminary maintenance scheduling instruction to the cloud. The cloud first obtains the latest real resource status of the optimal maintenance site to determine whether the site has been occupied.

[0156] The specific judgment logic is as follows: extract the current number of idle repair stations and the number of available spare parts in the latest real resource status of the corresponding optimal maintenance station from the cloud. If the number of idle repair stations is zero, or the number of available spare parts is less than the number required for vehicle repair, it is determined that the resources of the optimal maintenance station have been occupied.

[0157] Conversely, if the resource is not occupied, it is determined that it is not occupied. If the resource is not occupied, the maintenance resource of the station is locked to ensure that the current vehicle can enter the maintenance process first. If the resource is occupied, it means that multiple vehicles are pointing to the same maintenance resource at the same time. At this time, the cloud builds a scheduling priority list based on the maintenance requests uploaded by multiple vehicles.

[0158] When distributed vehicles concurrently restore network communication, in order to avoid multiple vehicles simultaneously determining that resources are not occupied and causing allocation conflicts due to communication link delays, the cloud-based judgment logic follows a global transaction mutex lock mechanism.

[0159] When multiple vehicles upload requests to the cloud within a very short time window and point to the same available maintenance resource, the system will trigger queuing contention protection, determine that the resource has entered a concurrent contention state, and import the requests of all relevant vehicles within the time window into the scheduling priority list for centralized comparison to ensure that there is no race deadlock on the collaborative scheduling timing link.

[0160] Here, the offline driving time is used to reflect the length of time the vehicle continues to drive without communication, the component wear value is used to reflect the current urgency of maintenance, and the first and second proportional coefficients are used to balance the contribution of the severity of wear and the length of waiting time to the priority.

[0161] The queuing priority score is obtained by adding the product of the component damage value and the first proportional coefficient, and the product of the network outage travel time and the second proportional coefficient.

[0162] This is used to incorporate the current level of vehicle wear and tear and the time the vehicle continues to drive after disconnecting from the network into the queuing decision, so as to avoid the uneven distribution of resources caused by sorting based on a single indicator.

[0163] To address the difference in dimensions and magnitude between component wear values ​​and network outage travel time, the system pre-maps both values ​​using the aforementioned minimax transformation method before performing the addition calculation. The range of values;

[0164] Among them, when the network outage travel time undergoes a maximum-minimum transformation, the global maximum value required for its denominator is fixed to the system's preset physical tolerance limit time constant. This is to avoid normalization divergence caused by the unbounded upper limit of the dependent variable, and the first and second proportional coefficients constitute the basis for the fusion of indicators of different dimensions.

[0165] The first proportional coefficient is determined based on the historical probability sensitivity of the damaged component to serious systemic failure, and the second proportional coefficient is determined based on the multiplier of the safety hazards caused by the vehicle's excessive driving time.

[0166] In a specific quantitative calibration example, the system sets the range of the first proportional coefficient to 0.6 to 0.8 and the range of the second proportional coefficient to 0.2 to 0.4, and the sum of the two is forced to be 1, so as to ensure the rigor of the multi-dimensional weighting logic and the reliability of the fusion of the underlying calculation data.

[0167] If the vehicle has the highest queuing priority score, the maintenance resources of the optimal maintenance station will be locked; otherwise, the alternative maintenance stations will be recalculated based on the global resource data stored in the cloud, and a redirection command will be output to redirect the vehicle to other available stations.

[0168] The specific logic for recalculating alternative repair stations is as follows: The cloud system removes the optimal maintenance station currently occupied by resource conflicts from the global candidate resource pool; Based on the vehicle's location coordinates uploaded at this time, it filters the road network status data again to select a new set of candidate stations that are within a preset distance range along the current driving trajectory heading angle and have available idle work positions and spare parts.

[0169] The aforementioned comprehensive scheduling index calculation module is invoked to calculate the spatial distance parameter, estimated arrival time parameter, and spare parts matching degree of the vehicle to each new candidate station. The weights are used to perform a weighted summation, and the new candidate station with the smallest comprehensive scheduling index is locked as the alternative maintenance station.

[0170] By establishing this suboptimal search feedback rule, we ensure that the redirected sites are optimally aligned with the vehicle's direction of travel, thus avoiding unpredictable optimization due to a lack of orientation constraint parameters.

[0171] This embodiment, through a transparent resource status verification mechanism and alternative site calculation based on preset suboptimal rules, can maintain a clear scheduling order and reduce resource conflicts when multiple vehicles request resources simultaneously.

[0172] This embodiment can generate a multi-vehicle collaborative scheduling strategy based on resource verification and dynamic priority sorting after network recovery, avoiding scheduling deadlock caused by multiple vehicles competing for the same maintenance resource at the same time, and ultimately solving the problem of the lack of a complete multi-vehicle collaborative anomaly handling mechanism in the prior art.

[0173] Based on the above embodiments, the execution result of the multi-vehicle scheduling strategy includes resource lock confirmation information and redirection instructions; the resource lock confirmation information is used to notify the vehicle-side target station that it has been locked, and the redirection instructions are used to notify the vehicle-side to change the maintenance target;

[0174] The offline cache flag in the scheduling status identifier indicates that the initial maintenance scheduling instruction has entered the local offline cache unit, the network synchronization flag indicates that the instruction has been uploaded to the cloud and entered the synchronization processing stage, and the resource occupancy flag indicates whether the target maintenance resource is currently occupied. After receiving the redirection instruction and status update information sent from the cloud, the network communication interface can synchronize and update the local cache and display content.

[0175] The vehicle-mounted display terminal is used to display the pre-decision guidance interface and the dispatch status interface on the vehicle side; the pre-decision guidance interface displays the component wear value, the optimal maintenance station and the estimated arrival time in the form of a list, so that the driver can directly view the current maintenance suggestions;

[0176] The scheduling status interface displays the network outage driving time, resource conflict handling results, and resource lock status in the form of logs, so that the vehicle can clearly understand the current scheduling results.

[0177] This embodiment adds a visual display on the vehicle side, which can ensure that the vehicle has consistent scheduling prompts before and after communication is restored. This embodiment can display the scheduling results based on the scheduling status identifier and the vehicle display terminal, avoiding the lack of feedback from the vehicle side on the scheduling process, and ultimately solving the problem of opaque vehicle-side information in the prior art.

[0178] In the application example, when the vehicle is in a communication-restricted environment, the system first caches the preliminary maintenance dispatch instructions locally and generates a network outage cache flag;

[0179] After communication is restored, the system will upload the command to the cloud, which will then verify the resource usage and return a resource lock confirmation message or a redirection command.

[0180] The vehicle-mounted display terminal updates the pre-decision guidance interface and the scheduling status interface accordingly, so as to keep the vehicle operation, maintenance resources and scheduling results consistent.

[0181] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A vehicle maintenance and repair operation management system, characterized in that, It includes a data acquisition module, a loss calculation module, a data distribution module, and a decision management module. The acquisition module is used to acquire multi-source sensor data during vehicle operation, calculate the data fluctuation amplitude based on continuous data within a preset time window, filter the data based on the data fluctuation amplitude, output key operating data, and transmit it to the loss calculation module. The loss calculation module is used to receive the key operating data, calculate the component loss value of the vehicle based on the preset loss change calculation model, output the component loss value, and transmit it to the decision management module. The distribution module is used to acquire global resource data, decompose the global resource data, generate surrounding maintenance resource data based on the acquired vehicle location coordinates, and distribute the surrounding maintenance resource data to the decision management module. The decision management module includes a local offline caching unit, a network status detection unit, and a network communication interface. It is used to receive the component wear value and cache the surrounding maintenance resource data. When the component wear value exceeds a preset physical safety limit, it generates a preliminary maintenance scheduling instruction based on the surrounding maintenance resource data and stores it in the local offline caching unit. When the network status detection unit detects network recovery, it uploads the preliminary maintenance scheduling instruction to the cloud through the network communication interface for resource conflict handling and outputs the maintenance scheduling result.

2. The vehicle maintenance and repair operation management system according to claim 1, characterized in that, The multi-source sensor data includes on-board diagnostic system data and chassis vibration data; The global resource data includes maintenance site data, spare parts resource data, and road network status data.

3. The vehicle maintenance and repair operation management system according to claim 1, characterized in that, The acquisition module generates the key operational data based on the multi-source sensor data, specifically including: A preset time window consisting of continuous sampling points is set, and continuous sampling time data of each sensor channel is extracted within the preset time window as the continuous data. Based on the time series change rate and local dispersion of the continuous data, features are extracted to obtain the data fluctuation amplitude; Arrange the non-zero elements in the data fluctuation range in descending order, and extract the values ​​corresponding to the preset quantiles as the filtering threshold. The data fluctuation amplitude is compared with the filtering threshold. If the data fluctuation amplitude is greater than the filtering threshold, the identifier corresponding to the multi-source sensor data is set to the reserved state. If the data fluctuation amplitude is less than or equal to the filtering threshold, the identifier corresponding to the multi-source sensor data is set to the filtering state, thus forming a data validity identifier. The key operational data is filtered out by aligning the multi-source sensor data with the data validity identifier by timestamp and multiplying them element by element.

4. The vehicle maintenance and repair operation management system according to claim 3, characterized in that, The loss calculation module generates the component loss value based on the key operating data, specifically including: Read the historical cumulative loss value as the current cumulative loss value, and obtain the preset prediction time interval; Based on the energy intensity of the effective elements in the key operational data, a nonlinear compression expression is performed to obtain the current loss increment index; The loss increment is calculated based on the current loss increment index, the preset wear correction coefficient, and the prediction time interval. The cumulative loss value at the current moment is added to the loss increment to generate the component loss value at the next moment; When the network is connected, the actual wear value sent from the cloud is received. The prediction deviation is obtained based on the actual wear value and the component wear value at the next moment. The wear correction coefficient is updated based on the current wear correction coefficient as a historical coefficient, a preset first weight, and a preset second weight.

5. The vehicle maintenance and repair operation management system according to claim 4, characterized in that, The distribution module generates surrounding maintenance resource data based on the global resource data, specifically including: Construct a resource association table that includes maintenance site dimensions, resource dimensions, and road network dimensions; The resource association table is decomposed into multiple data tables by splitting fields according to the maintenance site dimension, the resource dimension, and the road network dimension, and the site attribute data, spare parts attribute data, and road condition attribute data including road congestion coefficient are extracted. Obtain the continuous position coordinates of the vehicle, and calculate the heading angle of the driving trajectory based on the continuous position coordinates; Based on the continuous position coordinates, the driving trajectory heading angle, the station attribute data, the spare parts attribute data, and the road condition attribute data, surrounding maintenance resource data is generated within a preset distance and preset angle range, with the vehicle's current position as the starting point and the driving trajectory heading angle as the center line.

6. The vehicle maintenance and repair operation management system according to claim 5, characterized in that, The decision management module generates the preliminary maintenance scheduling instruction based on the component wear value and the surrounding maintenance resource data, specifically including: The component wear value is compared with the preset physical safety limit value. If the component wear value is less than or equal to the preset physical safety limit value, the current operating state of the vehicle is maintained. When the component wear value is greater than the preset physical safety limit value, the calculation of the comprehensive scheduling index for the candidate station is triggered. The candidate stations in the surrounding maintenance resource data are traversed, and the spatial distance parameter from the current location of the vehicle to the candidate station is calculated. The estimated arrival time parameter is calculated based on the estimated arrival time of the vehicle to the candidate station and the road congestion coefficient obtained from the road condition attribute data. The spare parts matching degree is determined based on the spare parts inventory quantity of the candidate station and the spare parts required by the vehicle. Based on the spatial distance parameter, the estimated arrival time parameter, and the spare parts matching degree, a comprehensive scheduling index is calculated, and the optimal maintenance site with the smallest comprehensive scheduling index is selected. A preliminary maintenance scheduling instruction containing the vehicle identifier, the key operating data, the component wear value, and the optimal maintenance site is generated and stored in the local offline cache unit.

7. The vehicle maintenance and repair operation management system according to claim 6, characterized in that, The decision management module generates a multi-vehicle dispatching strategy based on the preliminary maintenance dispatching instruction, specifically including: When the network status detection unit detects that the network has been restored, it uploads the preliminary maintenance scheduling instruction to the cloud. Obtain the latest real resource status of the optimal maintenance site and determine whether the resources of the optimal maintenance site are occupied; If the resources of the optimal maintenance station are not occupied, the maintenance resources of the optimal maintenance station will be locked directly for the vehicle. If the resources of the optimal maintenance site are occupied, a scheduling priority list is constructed based on the maintenance requests uploaded by multiple vehicles received from the cloud. Extract the network outage driving time of the vehicle since the generation of the initial maintenance scheduling instruction, call the preset first proportional coefficient and second proportional coefficient, and calculate the queuing priority score based on the component wear value, the network outage driving time and the preset first proportional coefficient and second proportional coefficient. If the queuing priority score is the highest, then the vehicle is locked with the maintenance resources of the optimal maintenance station; If the queuing priority score is not the highest, the alternative maintenance station will be recalculated based on the global resource data stored in the cloud, and a redirection instruction will be output.

8. The vehicle maintenance and repair operation management system according to claim 7, characterized in that, The execution result of the multi-vehicle scheduling strategy includes resource lock confirmation information and the redirection instruction; The decision management module generates a scheduling status identifier, which includes a network outage cache marker, a network synchronization marker, and a resource occupancy marker. When the preliminary maintenance scheduling instruction is stored in the local offline cache unit, the network outage cache marker is generated; The network communication interface is used to receive and parse the redirection instructions and status update information sent from the cloud.

9. A vehicle maintenance and repair operation management system according to claim 8, characterized in that, The vehicle maintenance and repair operation management system also includes an on-board display terminal, which is used to display a pre-decision guidance interface and a scheduling status interface. The pre-decision guidance interface is displayed in a list format, and the displayed content includes the component wear value, the optimal maintenance site, and the estimated arrival time. The scheduling status interface displays information in log format, including the network outage duration, resource conflict resolution results, and resource lock status.

Images