Road maintenance machinery spare part management method and device

Abstract

The invention provides a road maintenance machinery spare part management method and device. The method provided by the invention comprises the following steps: acquiring real-time spare part information of a spare part library of the road maintenance machinery, determining a part combination relationship among parts, and constructing a part sub-chain; determining a component combination relationship among the component sub-chains according to functions completed by the component sub-chains, and constructing a spare part link map; determining a required part number corresponding to each part based on a combination relationship in the spare part link graph; acquiring real-time operation state data of the road maintenance machinery; predicting a predicted fault of the road maintenance machine based on the real-time operation state data; predicting a target part demand number based on the prediction fault; according to the spare part link map and a target part, predicting the demand quantity of associated parts; summarizing to obtain the total part demand amount; and generating a spare part purchasing scheme based on the total part demand and the spare part link map. And the spare part management efficiency is improved.

Description

Technical Field

[0001] The present application relates to spare parts management in a parts library, and in particular to a method and device for managing spare parts of road maintenance machinery. Background Art

[0002] Track maintenance machinery is crucial equipment for railway infrastructure maintenance, and its operational stability directly impacts maintenance efficiency. Due to the complex structure and diverse component types of track maintenance machinery, spare parts inventory management is difficult. Traditional spare parts management relies on manual experience, making it difficult to keep track of spare parts inventory and demand in real time, leading to redundant or shortage spare parts. Some spare parts inventory management systems have incorporated computer technology, but this only enables real-time monitoring of spare parts status, i.e., synchronously updating spare parts inventory levels, outbound and inbound quantities. Spare parts are still purchased by management personnel based on experience, resulting in weak correlation between fault prediction and spare parts demand. Experience-based procurement cannot accurately predict impending faults, and procurement quantity predictions are inaccurate. Furthermore, procurement is often conducted solely based on human experience, without comprehensive consideration of factors such as part functional relevance and supply chain timeliness, which can easily lead to high procurement costs or maintenance delays. Summary of the Invention

[0003] A first aspect of the present application provides a method for managing spare parts of road maintenance machinery, the method comprising:

[0004] Acquire real-time spare parts information from a spare parts library of road maintenance machinery, wherein the real-time spare parts information includes parts and their corresponding part quantities;

[0005] Determine the part combination relationship between each part, use the parts as nodes, and use the part combination relationship as edges to connect the nodes corresponding to the corresponding parts to build a component sub-chain;

[0006] Determine the component combination relationship between each component sub-chain according to the functions completed by each component sub-chain, connect each component sub-chain based on the component combination relationship, and build a spare parts link map;

[0007] Determine the required parts quantity corresponding to each part based on the combination relationship in the spare parts link map;

[0008] Acquiring real-time operating status data of the road maintenance machinery;

[0009] Predicting a predicted fault of the road maintenance machine based on the real-time operating status data;

[0010] Predicting the required quantity of target parts based on the predicted failure;

[0011] Predicting the required quantity of associated parts based on the spare parts link map and the target parts;

[0012] Summarizing the required parts quantity, target parts demand quantity and associated parts demand quantity to obtain the total parts demand;

[0013] A spare parts procurement plan is generated based on the total parts demand and the spare parts link map.

[0014] A second aspect of the present application provides a road maintenance machinery spare parts management device, the device comprising:

[0015] An acquisition module, configured to acquire real-time spare parts information from a spare parts library of road maintenance machinery, wherein the real-time spare parts information includes parts and their corresponding part quantities;

[0016] A combination module is used to determine the part combination relationship between each part, use the parts as nodes, and use the part combination relationship as edges to connect the nodes corresponding to the corresponding parts to build a component sub-chain;

[0017] The combination module is further configured to determine a component combination relationship between each component sub-chain according to the functions performed by each component sub-chain, connect each component sub-chain based on the component combination relationship, and construct a spare parts link map;

[0018] A quantity estimation module, configured to determine the required quantity of each part based on the combination relationship in the spare parts link graph;

[0019] A monitoring module, used to obtain real-time operating status data of the road maintenance machinery;

[0020] A prediction module, configured to predict a predicted fault of the road maintenance machine based on the real-time operating status data;

[0021] The quantity estimation module is further configured to predict the required quantity of target parts based on the predicted fault; and predict the required quantity of associated parts based on the spare parts link map and the target parts;

[0022] A summarizing module is used to summarize the required parts quantity, target parts demand quantity and related parts demand quantity to obtain the total parts demand;

[0023] A procurement module is used to generate a spare parts procurement plan based on the total parts demand and the spare parts link map.

[0024] The road maintenance machinery spare parts management method and device provided in this application, first of all, compared with the conventional list-based management method, the method provided in this application combines parts into an organic whole in the form of complete components and functional modules, and uses a two-level (component level, functional module level) map method to manage parts, thereby improving the accuracy of subsequent searches and quantity updates. Secondly, when determining the purchase quantity of spare parts, the sources of the purchase quantity are determined from two perspectives. The first perspective is the real-time situation of the spare parts warehouse. Considering different reasons, the quantity of parts that are combined with each other does not match, which will lead to the inability to combine into the expected number of components or functional modules. Simply looking at the real-time inventory quantity of a part for spare parts is very inaccurate. This application realizes the first spare parts purchase quantity prediction from the level of components and functional modules in order to achieve quantity matching between related parts; from the second perspective, fault maintenance will lead to the use of spare parts. Taking into account the sufficient quantity during use and the replenishment of quantity after use, the functional modules and components required are screened according to the status of the fault, and then the quantity required to be purchased for the entire organic whole is determined. On the basis of the first aspect, the consideration unit of the spare parts strategy is changed from individual parts to component-level organic wholes, which improves the scientificity and efficiency of spare parts, thereby ensuring the effectiveness of maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 Flowchart of the road maintenance machinery spare parts management method provided in Example 1 of the present application;

[0026] Figure 2 This is a structural diagram of the road maintenance machinery spare parts management device provided in Example 2 of the present application. DETAILED DESCRIPTION

[0027] Exemplary embodiments are described in detail herein, with examples illustrated in the accompanying drawings. When the following description refers to the drawings, identical numerals in different drawings represent identical or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with this application.

[0028] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms "a," "the," and "the" used in this application are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0029] It should be understood that although the terms first, second, third, etc. may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0030] Specific embodiments are given below to introduce the technical solutions of the present application in detail.

[0031] Figure 1 This is a flowchart of the method for managing spare parts for road maintenance machinery provided in Example 1 of this application. Figure 1 The method provided in this embodiment may include:

[0032] S1: Acquire real-time spare parts information of a spare parts warehouse of road maintenance machinery, wherein the real-time spare parts information includes parts and their corresponding part quantities.

[0033] Spare parts for road maintenance machinery are usually stored in warehouses, and the information is updated as parts enter and exit the warehouse in real time. There are many ways to collect real-time spare parts information, including but not limited to: Internet of Things (IoT) device collection, deployment of RFID (radio frequency identification) tags or QR codes in the spare parts warehouse, and each part is bound to a unique identification code. Automatically record the entry, exit and inventory changes of parts through RFID readers or scanning devices, and upload them to the cloud database in real time. Warehouse Management System (WMS) docking: Establish a data interface with the company's warehouse management system (such as SAP WM, Oracle WMS) to synchronize inventory information regularly or in a triggered manner. Obtain data such as spare part name, specification model, current inventory quantity, storage location, purchase batch and expiration date through API calls. Manual entry on mobile terminals: Maintenance personnel scan the spare part barcode through a handheld terminal (PDA or mobile phone APP) to manually update the inventory status (such as loss, scrap). Support taking photos and uploading physical photos of spare parts, combined with AI image recognition technology to automatically match inventory records. Supply chain collaborative data: Connect with the supplier system to obtain information on spare parts in transit (such as the number of parts that have been ordered but not yet arrived and the estimated arrival time). Combined with blockchain technology to ensure that data cannot be tampered with and improve supply chain transparency. Sensor monitoring: Deploy weight sensors or optical counters in the storage area of ​​key spare parts (such as wearing parts) to monitor physical inventory changes in real time. For example: A pressure sensor is set in the hydraulic cylinder storage area to automatically trigger an early warning when the inventory decreases. As an optional embodiment, after the sensor collects the data, the data needs to be cleaned and verified. Specifically, the collected raw data is deduplicated, outliers are eliminated (such as negative inventory), and the format is standardized. Verify data consistency through the rule engine (such as "shipping quantity ≤ current inventory"). Output data example:

[0034]

[0035]

[0036] S2: Determine the part combination relationship between each part, use the parts as nodes, and use the part combination relationship as edges to connect the nodes corresponding to the corresponding parts to build a component sub-chain.

[0037] Road maintenance machinery is composed of numerous parts. These parts do not function independently but are combined into functional units through specific methods (such as mechanical connections, electrical coupling, or functional collaboration). Traditional spare parts management focuses solely on the inventory of individual parts, ignoring their combined relationships. This can lead to the omission of related spare parts during maintenance. This step uses graph structure modeling to visualize the physical and functional relationships between parts, providing a topological foundation for subsequent demand forecasting.

[0038] Specifically, the method of determining the parts combination relationship between each part, taking the parts as nodes, and taking the parts combination relationship as edges to connect the nodes corresponding to the corresponding parts, and constructing a component sub-chain includes: obtaining multiple structural diagrams of multiple types of road maintenance machinery and equipment; inputting all parts in the multiple structural diagrams into a machine learning model, and training a parts combination prediction model with the parts combination relationship in the structural diagram as output; obtaining parts in the real-time spare parts information, and inputting them into the trained parts combination prediction model to predict the combination relationship between each part in the spare parts library, wherein a part exists in multiple combination relationships, and the parts in each combination relationship are connected according to the combination relationship to form a component that independently performs a function; taking the combination relationship as an edge, connecting the nodes of the corresponding combination parts, wherein the edge value is the combination method, and the direction of the edge is the connection matching direction.

[0039] Specifically, a structural diagram is a diagram of the connections between parts of road maintenance machinery and equipment. Using mechanical structural diagrams, bills of materials (BOMs), or CAD assembly drawings, the connection methods between parts (such as bolt fixation, gear meshing, and hydraulic line docking) are analyzed to determine the physical connection relationships between the parts. Functional dependencies are determined based on the functions of the road maintenance machinery and equipment. If two parts work together to complete a function (such as "sealing"), an edge is established even if there is no physical contact. For example, hydraulic oil (node ​​C) and a sealing ring (node ​​D) are associated through "sealing functional dependency" (edge ​​attributes). The specific connection situation is determined based on the physical connection relationship and functional dependency.

[0040] When constructing a component subchain, nodes represent individual parts and store attributes such as part ID, name, specifications, and part quantity. Edges represent combinational relationships and store attributes such as connection type (e.g., mechanical connection, electrical coupling, functional collaboration); directionality (e.g., hydraulic oil flow determines edge direction); and strength weight (e.g., the tightness of bolted connections influences requirement relevance). A component subchain is a set of nodes tightly connected by edges, corresponding to a component that independently performs a function (e.g., a "brake module" or "hydraulic pump unit"). Multiple links within a component subchain must be replaced simultaneously (e.g., a seal and its matching flange). A component subchain contains multiple nodes, or multiple parts. Each component subchain may perform the same or different functions. Component subchains that perform the same function may not contain identical parts, and their combinations may also differ. Component subchains that perform different functions may contain completely different parts, or they may share some nodes. For quantities that share identical nodes, each node stores the same part quantity attribute.

[0041] S3: Determine the component combination relationship between each component sub-chain according to the function completed by each component sub-chain, connect each component sub-chain based on the component combination relationship, and construct a spare parts link map.

[0042] This step elevates the physical layer (component sub-chain) to the functional layer (spare parts link map) through functional topology aggregation, realizing two-level functional hierarchical parts management, visual transmission of fault impact range, and collaborative calculation of spare parts requirements across components.

[0043] Specifically, the component combination relationship between each component sub-chain is determined according to the function completed by each component sub-chain, and each component sub-chain is connected based on the component combination relationship, including: for each component sub-chain, determining the function completed after the parts in the component sub-chain are combined as the representative function of the component sub-chain; determining the representative node in the component sub-chain according to the representative function and the edge in the component sub-chain; determining the component combination relationship between each component sub-chain according to the representative function; and connecting the representative nodes of the corresponding component sub-chains with the component combination relationship as the edge, wherein the value of the edge is the combination method, and the direction of the edge is the connection matching direction.

[0044] Among them, the representative function refers to the most core function among the functions completed by the combination, and the other functions in the functions completed by the combination are the supporting functions of the most core function. The representative node is the most core node. The connection mode and dependency direction between the parts corresponding to each node can be judged according to the attributes of the edge, and the status of each node can be determined according to the connection mode and dependency direction. Among them, since the spare parts link map provided by this application is two-level, the representative nodes also have two levels, that is, there is a representative node in each component sub-chain, and there is also a representative node in the functional module after the component sub-chain is combined. The representative nodes of the two levels are not exactly the same.

[0045] S4: Determine the required parts quantity corresponding to each part based on the combination relationship in the spare parts link map.

[0046] The purpose of step S4 is to adapt the existing inventory quantity in the spare parts library so that the functional module chain and the component sub-chain can have corresponding numbers. The required quantity is determined according to the two levels of the spare parts link map. As an optional embodiment, the required number of parts corresponding to each part is determined based on the combination relationship in the spare parts link map, including: determining the functional module chain after the component sub-chain and the component sub-chain are combined according to the combination relationship in the spare parts link map; determining the representative nodes of the component sub-chain and the functional module chain respectively according to the completed functions and edge attributes; determining the first number of parts corresponding to the representative node of the functional module chain in the spare parts library according to the number of parts in the real-time spare parts information; determining the second required number of parts corresponding to the representative node of the component sub-chain based on the first number; determining the second required number of parts corresponding to the non-representative node based on the second required number of the representative node and the component sub-chain; arranging the second required number of each part in the spare parts link map in sequence to obtain the required number of parts corresponding to each part.

[0047] Specifically, there are two levels of combinations in the spare parts link map. The first is the functional module combination. For a functional module combination, the number of pairs of repeated wholes that can be combined is determined by the number of parts corresponding to the representative nodes. For example, for a functional module, there are 20 part A, so 20 sets of wholes need to be combined, that is, 20 sets of parts of the functional module other than the representative nodes are required. For node B, if 2 parts are required to combine a set, then 40 parts corresponding to node B are required, and so on, to obtain the required number of parts for non-representative nodes; when it comes to node C, node C is the representative point of a component sub-chain in the functional module. At this time, determine how many parts corresponding to the component sub-chain where node C is located are required for 1 set of functional module combination, and determine the quantity requirement of node C based on the required number. Then, similar to the method of calculating the number of nodes in the functional module, take node C as the representative node and calculate the quantity requirement of non-representative nodes in the component sub-chain where node C is located.

[0048] The core purpose of this step is to analyze the combination relationships between spare parts, dynamically adapt the existing inventory in the spare parts library to the actual demand of functional module chains and component subchains, and ultimately calculate the specific required quantity of each part. The system parses the combination relationships recorded in the spare parts link graph (such as directed edges, parent-child nodes, etc.) and identifies two types of key chains: component subchains: subassembly units formed by hierarchical combination of basic parts (such as a fastening unit consisting of a bolt and a gasket), which are connected by multiple parts; and functional module chains: complete functional modules formed by further combination of multiple component subchains (such as a drive module containing a motor, gearbox, and housing). Based on the functional completeness of the chain and the attributes of the edges (such as weights and dependencies), the system assigns a representative node to each chain: for functional module chains, the representative node is typically the core component directly associated with the final function (such as the motor in the drive module); for component subchains, the representative node is the key part of the subassembly (such as the bolt in the fastening unit). The first quantity can be obtained by querying the real-time spare parts information database to obtain the current inventory quantity (i.e., the first quantity) of the parts corresponding to the functional module chain representative node. For example, if the motor inventory is 5 units, the first quantity is 5. The demand of the component sub-chain representative node is directly mapped according to the first quantity. For example, if each drive module requires 1 motor and the motor inventory is 5, the second demand quantity corresponding to the motor is 5. Based on the combination ratio of the component sub-chain (such as 1 bolt requires 2 gaskets), the demand quantity of the non-representative node is calculated step by step by traversing the topological relationship of the sub-chain. For example, if the bolt demand is 5, the gasket demand is 5×2=10. The second demand quantity of each part is arranged in the hierarchical order of the spare parts link map (such as functional module → sub-chain → basic parts), and a structured demand list is generated, and the final output demand list is: pump housing (3), bearing (6), O-ring (3), gasket (9).

[0049] S5: Acquire real-time operating status data of the road maintenance machinery.

[0050] In step S5, the system acquires real-time operating status data of the road maintenance machinery to facilitate subsequent spare parts demand forecasting, fault diagnosis, and maintenance decision-making. The key to this step is the real-time collection of the machinery's operating parameters, operating conditions, and equipment health status, ensuring data freshness and accuracy.

[0051] The real-time operating status data can be obtained through at least one of the following methods: On-board sensor data: Various sensors installed on the road maintenance machinery (such as vibration sensors, temperature sensors, pressure sensors, speed sensors, etc.) monitor the operating status of the machinery in real time. For example: engine speed, oil temperature, and oil pressure; hydraulic system pressure and flow; transmission component vibration amplitude and temperature. Equipment control unit (ECU / PLC) data: Operating parameters are read from the road maintenance machinery's electronic control unit (ECU) or programmable logic controller (PLC), such as: operating mode (such as milling mode, compaction mode); load current and voltage; fault codes (such as overload alarm, oil line blockage). Remote monitoring system (IoT) data: If the road maintenance machinery is equipped with an Internet of Things (IoT) module, data can be uploaded to the cloud server in real time via wireless communication (such as 4G / 5G, LoRa), including: GPS positioning information (operating location, movement trajectory); accumulated operating time (used to calculate component wear); and real-time alarm information (such as abnormal shutdown and overtemperature alarm). Manual data input: In some scenarios, operators can manually enter the machine's operating conditions through mobile terminals (such as tablets and handheld devices), such as: current operation tasks (such as rail grinding and ballast cleaning); environmental conditions (such as temperature, humidity, and dust level).

[0052] After obtaining the real-time operating status data, the method also includes preprocessing and standardizing the real-time operating status data. Due to the diverse sources of data, the system can preprocess the collected raw data to ensure the consistency and availability of the data, specifically including: data cleaning: eliminating abnormal values ​​(such as sudden increases in data caused by sensor failure) and filling missing values ​​(such as using the sliding average method for interpolation). Data normalization: standardizing data of different dimensions (such as temperature unit ℃, pressure unit MPa) (such as scaling to the range of 0 to 1). Data fusion: performing correlation analysis on multi-source data, such as combining vibration data with speed data to determine the health status of the bearing. The collected data can be stored in a local cache (such as an on-board data recorder) or a cloud database, and indexed according to timestamps for subsequent analysis. The system can set the data sampling frequency (such as once per second or once per minute) to meet the monitoring needs of different components (such as high-speed rotating components require a higher sampling rate).

[0053] S6: Predicting a predicted fault of the road maintenance machine based on the real-time operating status data.

[0054] In step S6, the system predicts potential faults of the road maintenance machinery based on the real-time operating status data (e.g., sensor data, equipment control parameters, and operating condition information). This step utilizes a pre-trained fault prediction model, combined with real-time monitoring data, to proactively identify potential faults and assess their type, probability, and severity, enabling preventive maintenance measures to reduce unplanned downtime.

[0055] As an optional embodiment, the predicting of the predicted fault of the road maintenance machinery based on the real-time operating status data includes: collecting the real-time operating status data of the road maintenance machinery; inputting the real-time operating status data into a pre-trained fault prediction model; and predicting the fault type, fault probability and fault severity of the road maintenance machinery through the fault prediction model.

[0056] The fault prediction model can be trained using at least one of the following machine learning or deep learning algorithms, such as a supervised learning model (such as random forest, support vector machine, XGBoost), or a time series prediction model (such as LSTM, Transformer). Specifically, it is trained based on historical fault data (such as fault logs, maintenance records) and corresponding operating status data (such as vibration, temperature, pressure, etc.) to establish a mapping relationship of "operating status → fault type". Input features may include: sensor time series data, equipment operating time, load change trend, etc. The output is the fault classification (such as bearing wear, hydraulic leakage) and the probability of occurrence (range 0 to 1).

[0057] Model training data comes from historical failure case libraries (such as the failure records of a certain model of tamping vehicle in the past five years), simulation data (simulating typical failure modes through digital twins), and cross-domain data of similar equipment (transfer learning).

[0058] After the model is trained, the fault prediction based on real-time operating status data includes the following steps: obtaining the real-time operating status data from step S5 (e.g., vibration = 6.3m / s2, oil temperature = 82°C, current fluctuation = ±5%); performing feature extraction on the real-time operating status data based on the fault prediction model (e.g., calculating the FFT spectrum of the vibration signal). The output of the fault prediction model includes at least the following three types of information: fault type: classification result (e.g., "hydraulic pump seal failure" or "bearing fatigue crack"); probability of occurrence: numerical confidence level (e.g., 0.85 represents 85% probability of occurrence); severity: graded label (e.g., "mild," "moderate," or "urgent").

[0059] As an optional embodiment, after the fault prediction model is trained, it also includes an online learning step to continuously optimize the model according to newly occurring fault cases (such as incremental training).

[0060] S7: Predicting the required quantity of target parts based on the predicted failure.

[0061] In step S7, the system predicts the required quantity of the target part based on the predicted fault (including fault type, probability, and severity). This step analyzes the scope of the fault impact and combines it with the part relationships in the spare parts link map to dynamically calculate the parts that may need to be replaced and the required quantity, thereby optimizing spare parts inventory management and improving maintenance efficiency.

[0062] As a preferred embodiment, the predicted fault includes at least: fault type, fault probability and fault severity; the predicted target parts demand quantity based on the predicted fault includes: determining the fault center location point according to the fault type and fault severity; determining multiple candidate fault nodes in the spare parts link map based on the fault center location point; screening target parts to be used according to the functions when the predicted fault is not a fault; calculating the replacement probability of the target part according to the fault probability; and calculating the target parts demand quantity according to the replacement probability.

[0063] Failures can occur at central and secondary locations. The central location is where the failure initially occurs, while the secondary locations are affected by the central location failure. At the location where the failure occurs later, both the central and secondary locations fail, resulting in the failure predicted by the model. The central location corresponds to a part, such as a screw. If a replacement is required, the node where the screw is located is found in the spare parts link map as a candidate failure node. However, the screw part is required in multiple chains. To accurately locate the part, the function chain in the chain where the candidate failure node is located is determined based on the predicted function when the failure is not occurring. The target part to be used is then filtered from these function chains.

[0064] Specifically, the fault center point refers to the core component directly affected by the fault. Preferably, this can be determined using fault type mapping, based on a predefined fault-component mapping table (e.g., "hydraulic leakage → hydraulic pump seal"). The mapping results of the mapping table are weighted according to severity. If the fault severity is "mild," the center point may only involve a single component (e.g., an O-ring); if it is "urgent," the center point may cover the entire functional module (e.g., the hydraulic pump assembly). For example, if the predicted fault is "bearing overheating (probability 75%, severity = moderate)," its center point is determined to be the "drive shaft bearing seat" through a knowledge base query.

[0065] Based on the fault's central location, associated candidate fault nodes are extracted from the spare parts link library. The spare parts link graph is traversed to identify nodes directly connected to the central point (e.g., adjacent components of a bearing housing: bearings, seals, and mounting bolts). If the central point belongs to a functional module (e.g., a drive unit), all subcomponents within that module are included as candidates. Based on the predicted functional requirements if the fault does not occur, candidate nodes are selected for replacement. This selection process first involves a functional necessity check. Parts whose damage would cause complete functional failure (e.g., a broken bearing causing a transmission stop) are marked as such. Parts that only affect performance but do not disrupt functionality (e.g., a minor seal leak) are not considered. Second, cost-effectiveness filtering can be performed: parts with a repair cost below a threshold (e.g., requiring only cleaning, not replacement) are excluded. For example, in the case of a bearing overheating fault, the bearing (core component) and seal (deformed) are selected, while mounting bolts (undamaged) are excluded.

[0066] For each target part, the replacement probability (P_replace) is calculated based on the failure probability (P_fault). Parts with different replacement levels are assigned replacement probabilities. If the part is the primary cause of the failure (e.g., a bearing), P_replace = P_fault (e.g., 75%). If the part requires preventive replacement due to a correlation (e.g., a seal), the replacement probability is calculated based on the product of the replacement probability of the primary cause of the failure and the coupling coefficient. The final required quantity is generated based on the replacement probability and spare parts inventory strategy. If P_replace exceeds a threshold (e.g., 50%), the additional quantity is increased by ceil(P_replace / threshold), where ceil is a round-up function.

[0067] S8: Predicting the required quantity of associated parts based on the spare parts link map and the target parts.

[0068] In step S8, the system predicts the required quantity of associated parts based on the spare parts link map and the target part. This step analyzes the topological relationship of the target part in the map and combines it with a functional compatibility assessment to determine the associated parts that need to be replaced and their required quantity, thereby achieving a more comprehensive spare parts demand forecast.

[0069] The method of predicting the required quantity of associated parts based on the spare parts link map and the target part includes: determining a target node position of the target part in the spare parts link map; determining a component sub-chain having a connection relationship with the target node position to obtain a plurality of candidate component sub-chains; calculating the fitness between the functions corresponding to the plurality of candidate component sub-chains and the faulty road maintenance machinery; screening the target component sub-chain based on the fitness; calculating the replacement probability of each part in the target component sub-chain based on the replacement probability of the target part; and calculating the required quantity of associated parts for each part in the target component sub-chain based on the replacement probability.

[0070] Generally speaking, the failure of a part will lead to secondary failure of the combined parts, and when the part is replaced, it is also replaced completely in the form of components. At this time, simply adjusting the inventory of a part is not conducive to the maintenance efficiency of the component. The method provided in this application finds the associated nodes according to the connection relationship in the graph, adjusts the required quantity, and always keeps the number of parts in the spare parts library adapted.

[0071] Specifically, locate the corresponding node (target node) in the spare parts link map based on the unique identifier of the target part; identify the level to which the target node belongs (such as basic parts, subassemblies, functional modules). For example, if the target part is a "hydraulic pump spindle bearing", locate the node "Bearing_001" in the spare parts link map, which belongs to the "hydraulic pump drive unit" subassembly. Based on the directed edge relationship of the graph, edges with direct parent-child relationships (such as the "spindle assembly" to which the bearing belongs), same-level dependency relationships (such as the "seal component" that cooperates with the bearing), and functional coupling relationships (such as the "coupling unit" that shares the same power source) with the target node are extracted, and each candidate sub-chain is labeled. A multidimensional evaluation model is performed based on functional relevance, physical connection strength, fault propagation risk, and difficulty of maintenance and disassembly. The weighted sum of each evaluation result is calculated based on the model evaluation to screen sub-chains that meet the conditions; a probability propagation model is used to calculate the replacement probability of each part in the sub-chain, and the connection coefficient is determined according to different connection relationships. The replacement probability of each part in the sub-chain is calculated based on the product between the connection coefficient and the corresponding replacement probability of the target node.

[0072] As another optional embodiment, key nodes in the spare parts link map are identified; wear coefficients of the key nodes are determined based on historical maintenance data; and the required number of parts is calculated based on the wear coefficients. Associated parts directly connected to the target part are identified in the spare parts link map; the probability of associated damage of the associated parts is determined based on the part combination relationship; and the required number of associated parts is calculated based on the associated damage probability.

[0073] S9: Summarize the required parts quantity, target parts demand quantity and associated parts demand quantity to obtain the total parts demand quantity.

[0074] In step S9, the system aggregates the required parts quantity, target parts quantity, and associated parts quantity to generate the final total parts demand. This step uses multi-level sorting and merging optimization to ensure that the generated parts demand list is both complete and consistent with actual maintenance priorities, facilitating inventory management and procurement decisions.

[0075] As a preferred embodiment, the required part quantity, target part demand quantity and associated part demand quantity are arranged in sequence to form a first part demand sequence, wherein each sub-element in the first part demand sequence includes the component sub-chain where the part is located, the part name and the required quantity; the importance of each sub-element is calculated according to the component sub-chain where the part is located; the sorting of each sub-element is adjusted based on the importance to obtain a second part demand sequence; the demand quantities of sub-elements with the same part name in the second part demand sequence are merged, and the component sub-chain information of the parts after the merged quantity is merged to obtain a third part demand sequence.

[0076] Specifically, a sequence is generated according to the connection order in the spare parts link map, and then the order is modified according to the importance. The spare parts procurement plan generated according to this order is also processed in a way that the important spare parts are purchased first, which improves the efficiency of spare parts. At the same time, the parts are prepared according to the importance of the parts they need to use to ensure that the maintenance is carried out normally. The required number of parts is obtained from step S4, the required number of target parts is obtained from step S7, and the required number of related parts is obtained from step S8 for aggregation. When merging, the quantity merging is to accumulate the required quantities of parts with the same name, and other information is merged: the sub-chain information of the components to be merged is combined with separators.

[0077] S10: Generate a spare parts procurement plan based on the total parts demand and the spare parts link map.

[0078] In step S10, the system generates an optimal spare parts procurement plan based on the total parts demand and the spare parts linkage map. This step intelligently matches inventory gaps with supplier resources and introduces a time-optimization algorithm to ensure that urgent maintenance needs are met while controlling costs.

[0079] As a preferred embodiment, based on the third parts demand sequence and the real-time spare parts information in the spare parts warehouse, the part difference corresponding to each sub-element is determined to generate a difference sequence; and a spare parts procurement plan is generated based on the difference sequence. Specifically, the nearest purchaser corresponding to each difference element in the difference sequence is determined, where the number of parts available in the nearest purchaser is greater than or equal to the difference element; the delivery time required for the nearest purchaser to reach the spare parts warehouse is calculated; the order of each difference element in the difference sequence is changed based on the delivery time to generate a procurement sequence; and the spare parts procurement plan is generated based on the order of the procurement sequence and the nearest purchaser corresponding to each element.

[0080] Specifically, traverse each sub-element in the third part demand sequence, extract the part name (such as "bearing_6205") and the total demand (such as 5 pieces), query the real-time spare parts library to obtain the current inventory (such as 2 pieces) and the safety stock threshold (such as 1 piece), and for each difference element, screen suppliers that meet the requirements. The inventory of the selected supplier must be ≥ the purchase difference. Only in this way can the purchase demand be met at one time from one supplier. Calculate the weighted sum based on the delivery time, purchase unit price, and historical delivery coordination to obtain the supplier's score, and select the supplier with the highest score, where the weight is a dynamic weight.

[0081] As an optional embodiment, the method further includes: regularly updating the spare parts link map; correcting the parts combination relationship and component combination relationship according to actual maintenance records; and optimizing subsequent parts demand forecasts based on the corrected spare parts link map.

[0082] In addition, a spare parts inventory early warning mechanism is established; when the number of parts in the real-time spare parts information is lower than the safety threshold, the spare parts procurement process is automatically triggered.

[0083] Corresponding to the aforementioned embodiment of a method for managing spare parts of road maintenance machinery, the present application also provides an embodiment of a device for managing spare parts of road maintenance machinery.

[0084] Figure 2 This is a schematic diagram of the structure of the road maintenance machinery spare parts management device provided in Example 2 of this application. Figure 2 , the device provided in this embodiment includes:

[0085] An acquisition module, configured to acquire real-time spare parts information from a spare parts library of road maintenance machinery, wherein the real-time spare parts information includes parts and their corresponding part quantities;

[0086] A combination module is used to determine the part combination relationship between each part, use the parts as nodes, and use the part combination relationship as edges to connect the nodes corresponding to the corresponding parts to build a component sub-chain;

[0087] The combination module is further configured to determine a component combination relationship between each component sub-chain according to the functions performed by each component sub-chain, connect each component sub-chain based on the component combination relationship, and construct a spare parts link map;

[0088] A quantity estimation module, configured to determine the required quantity of each part based on the combination relationship in the spare parts link graph;

[0089] A monitoring module, used to obtain real-time operating status data of the road maintenance machinery;

[0090] A prediction module, configured to predict a predicted fault of the road maintenance machine based on the real-time operating status data;

[0091] The quantity estimation module is further configured to predict the required quantity of target parts based on the predicted fault; and predict the required quantity of associated parts based on the spare parts link map and the target parts;

[0092] A summarizing module is used to summarize the required parts quantity, target parts demand quantity and related parts demand quantity to obtain the total parts demand;

[0093] A procurement module is used to generate a spare parts procurement plan based on the total parts demand and the spare parts link map.

[0094] The device of this embodiment can be used to perform Figure 1 The steps, specific implementation principles and implementation processes of the method embodiment shown are similar and will not be repeated here.

[0095] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.

[0096] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial description of the method embodiments. The device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application scheme. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0097] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for managing spare parts of road maintenance machinery, characterized in that: The method comprises: Acquire real-time spare parts information from a spare parts library of road maintenance machinery, wherein the real-time spare parts information includes parts and their corresponding part quantities; Determine the part combination relationship between each part, use the parts as nodes, and use the part combination relationship as edges to connect the nodes corresponding to the corresponding parts to build a component sub-chain; Determine the component combination relationship between each component sub-chain according to the functions completed by each component sub-chain, connect each component sub-chain based on the component combination relationship, and build a spare parts link map; Determine the required parts quantity corresponding to each part based on the combination relationship in the spare parts link map; Acquiring real-time operating status data of the road maintenance machinery; Predicting a predicted fault of the road maintenance machine based on the real-time operating status data; Predicting the required quantity of target parts based on the predicted failure; Predicting the required quantity of associated parts based on the spare parts link map and the target parts; Summarizing the required parts quantity, target parts demand quantity and associated parts demand quantity to obtain the total parts demand; A spare parts procurement plan is generated based on the total parts demand and the spare parts link map.

2. The method according to claim 1, characterized in that The determining of the parts combination relationship between the parts, taking the parts as nodes, and using the parts combination relationship as edges to connect the nodes corresponding to the parts, to construct a component sub-chain, includes: Obtain multiple structural drawings of multiple types of road maintenance machinery and equipment; Input all parts in the multiple structural diagrams into a machine learning model, and use the part combination relationship in the structural diagram as output to train a part combination prediction model; Obtaining parts from the real-time spare parts information and inputting them into a trained parts combination prediction model to predict the combination relationships between the parts in the spare parts library, wherein a part exists in multiple combination relationships, and the parts in each combination relationship are connected according to the combination relationship to form a component that independently completes a function; Use the combination relationship as the edge to connect the nodes of the corresponding combination parts, where the edge value is the combination method and the direction of the edge is the connection matching direction.

3. The method according to claim 1, characterized in that The determining of the component combination relationship between the component sub-chains according to the functions performed by the component sub-chains, and connecting the component sub-chains based on the component combination relationship, includes: For each component sub-chain, the function completed by the combination of parts in the component sub-chain is determined as the representative function of the component sub-chain; Determining a representative node in the component sub-chain according to the representative function and the edges in the component sub-chain; Determining component combination relationships between various component sub-chains based on the representative functions; The component combination relationship is used as an edge to connect the representative nodes of the corresponding component sub-chain, wherein the edge value is the combination method and the direction of the edge is the direction of the connection matching.

4. The method according to claim 1, wherein The determining of the required number of parts corresponding to each part based on the combination relationship in the spare parts link map includes: Determining the component sub-chain and the functional module chain after the component sub-chain is combined according to the combination relationship in the spare parts link map; Determining representative nodes of the component sub-chain and the functional module chain respectively according to the completed functions and edge attributes; Determine a first quantity of parts corresponding to a representative node of the functional module chain in a spare parts library according to the quantity of parts in the real-time spare parts information; Determine a second required quantity of the part corresponding to the representative node of the component sub-chain based on the first quantity; Determining the second required quantity of the parts corresponding to the non-representative nodes based on the second required quantity of the representative node and the component sub-chain; The second required quantities of each part in the spare parts link map are arranged in sequence to obtain the required parts quantity corresponding to each part.

5. The method according to claim 1, wherein The predicting of the predicted fault of the road maintenance machine based on the real-time operating status data includes: Collecting real-time operating status data of the road maintenance machinery; Inputting the real-time operating status data into a pre-trained fault prediction model; The fault prediction model is used to predict the fault type, fault probability and fault severity of the road maintenance machinery.

6. The method according to claim 1, characterized in that The predicted fault includes at least: fault type, fault probability and fault severity; The prediction of target parts demand quantity based on the predicted fault includes: Determine the fault center location point according to the fault type and fault severity; Determine multiple candidate fault nodes in the spare parts link map based on the fault center location point; Filter target parts to be used according to the functions when the predicted fault is not a fault; Calculating a replacement probability of the target part according to the failure occurrence probability; The target part requirement quantity is calculated according to the replacement probability.

7. The method according to claim 1, characterized in that The predicting of the required quantity of associated parts according to the spare parts link map and the target parts includes: Determine a target node position of the target part in the spare parts link graph; Determine a component sub-chain that has a connection relationship with the target node position, and obtain multiple candidate component sub-chains; Calculate the compatibility between the functions corresponding to multiple candidate component sub-chains and the faulty road maintenance machinery; screening a target component sub-chain based on the fitness; Calculating the replacement probability of each part in the target component sub-chain according to the replacement probability of the target part; The associated parts requirement quantity of each part in the target component sub-chain is calculated according to the replacement probability.

8. The method according to claim 1, characterized in that The step of aggregating the required parts quantity, the target parts demand quantity, and the associated parts demand quantity to obtain a total parts demand quantity, and generating a spare parts procurement plan based on the total parts demand quantity and the spare parts link map includes: Arranging the required part quantity, target part required quantity, and associated part required quantity in sequence to form a first part requirement sequence, wherein each sub-element in the first part requirement sequence includes the component sub-chain where the part is located, the part name, and the required quantity; Calculate the importance of each sub-element based on the sub-chain of the component where the part is located; Adjusting the order of each sub-element based on the importance to obtain a second parts requirement sequence; Merging the required quantities of sub-elements with the same part name in the second part requirement sequence, and merging the component sub-chain information of the parts with the merged quantities to obtain a third part requirement sequence; Determine the part difference corresponding to each sub-element according to the third part demand sequence and the real-time spare parts information in the spare parts library, and generate a difference sequence; A spare parts procurement plan is generated based on the difference sequence.

9. The method according to claim 8, characterized in that Generating a spare parts procurement plan based on the difference sequence includes: Determining the nearest purchaser corresponding to each difference element in the difference sequence, wherein the quantity of parts available in the nearest purchaser is greater than or equal to the difference element; Calculate the delivery time from the nearest purchaser to the spare parts warehouse; Changing the order of each difference element in the difference sequence based on the delivery time efficiency to generate a procurement sequence; A spare parts procurement plan is generated based on the order of the procurement sequence and the nearest purchaser corresponding to each element.

10. A spare parts management device for road maintenance machinery, characterized in that: The device comprises: An acquisition module, configured to acquire real-time spare parts information from a spare parts library of road maintenance machinery, wherein the real-time spare parts information includes parts and their corresponding part quantities; A combination module is used to determine the part combination relationship between each part, use the parts as nodes, and use the part combination relationship as edges to connect the nodes corresponding to the corresponding parts to build a component sub-chain; The combination module is further configured to determine a component combination relationship between each component sub-chain according to the functions performed by each component sub-chain, connect each component sub-chain based on the component combination relationship, and construct a spare parts link map; A quantity estimation module, configured to determine the required quantity of each part based on the combination relationship in the spare parts link graph; A monitoring module, used to obtain real-time operating status data of the road maintenance machinery; A prediction module, configured to predict a predicted fault of the road maintenance machine based on the real-time operating status data; The quantity estimation module is further configured to predict the required quantity of target parts based on the predicted fault; and predict the required quantity of associated parts based on the spare parts link map and the target parts; A summarizing module is used to summarize the required parts quantity, target parts demand quantity and related parts demand quantity to obtain the total parts demand; A procurement module is used to generate a spare parts procurement plan based on the total parts demand and the spare parts link map.

Images