Equipment maintenance work order dispatching method and route planning system for multiple campuses

Abstract

This application provides a method and path planning system for dispatching equipment maintenance work orders across multiple campuses, belonging to the field of multi-campus management technology. By constructing a digital twin model of a six-level topology—"campus-building-space-equipment-parts-maintenance personnel"—and decomposing maintenance work orders into feature vectors representing skill requirements, parts requirements, and energy consumption sensitivity, an initial dispatch plan and path planning are generated based on the topological relationships and feature vectors. During execution, the dispatch plan and path planning are dynamically adjusted based on real-time feedback from the equipment monitoring system. This invention achieves integrated intelligent decision-making and dynamic optimization of parts, personnel, paths, and energy consumption, effectively solving the problems of resource coordination difficulties and response delays in multi-campus maintenance, and significantly improving maintenance efficiency and resource utilization.

Description

Technical Field

[0001] This application relates to the field of multi-park management technology, and in particular to a method and path planning system for dispatching equipment maintenance work orders for multiple parks. Background Technology

[0002] As enterprises expand, managing equipment maintenance across multiple geographically dispersed campuses has become a complex challenge. Current technologies often rely on manual experience or simple polling rules for work order assignment, and path planning typically only considers the shortest distance or time. This results in "data silos" between various management systems (such as equipment management, energy management, and inventory management), lacking effective collaboration mechanisms.

[0003] Specifically, existing technologies have the following main problems: First, work order assignment is disconnected from spare parts inventory, often resulting in maintenance personnel making wasted trips due to spare parts shortages, leading to low efficiency; second, route planning is static and cannot be dynamically adjusted according to real-time traffic flow in the park, microgrid energy consumption peaks, and sudden new work orders, resulting in response delays and resource waste; finally, maintenance decisions lack a comprehensive consideration of the impact on energy consumption, making it difficult to support the company's green and energy-saving goals.

[0004] Therefore, there is an urgent need in this field for a method and system that can break down system barriers and achieve integrated intelligent decision-making and dynamic optimization of inventory, personnel, routes and energy consumption, so as to improve the overall efficiency and economy of equipment operation and maintenance in multiple parks. Summary of the Invention

[0005] To address the aforementioned technical issues, this application provides a method for dispatching equipment maintenance work orders and a path planning system for multiple campuses.

[0006] A first aspect of this application provides a method for dispatching equipment maintenance work orders across multiple campuses, including the following steps:

[0007] S100: Construct a multi-campus digital twin model. The digital twin model is configured with a six-level topology of "campus-building-space-equipment-parts-maintenance personnel" and dynamic attribute vectors are written for each level of the topology.

[0008] S200: Receive maintenance work orders, extract key features based on the maintenance work orders, and break down the maintenance work orders into feature vectors including skill requirements, parts requirements, and energy consumption sensitivity based on the key features;

[0009] S300: Generates an initial dispatch scheme based on a six-level topology and feature vectors;

[0010] S400: Based on the initial dispatch plan, perform preliminary route planning for maintenance personnel;

[0011] S500: During the execution of the initial route plan by the maintenance personnel, the initial dispatch plan is dynamically adjusted based on the parameters fed back in real time by the equipment monitoring system to obtain an optimized dispatch plan;

[0012] S600: The optimized path is obtained by adjusting the initial path planning based on the optimized dispatching scheme;

[0013] S700: Maintenance personnel perform maintenance on the equipment based on optimized routes and optimized dispatching schemes.

[0014] This application constructs a digital twin model of a six-level topology—"park-building-space-equipment-parts-maintenance personnel"—providing a unified and precise data foundation for the entire operation and maintenance process, achieving deep integration of the physical and information worlds. By breaking down work orders into multi-dimensional feature vectors such as skills, parts, and energy consumption, it achieves refined analysis of operation and maintenance requirements. The core innovation of this method lies in designing the optimization of the dispatch scheme and the optimization of the path planning as two dynamically feedback-enabled stages. This allows the system to continuously self-adjust based on real-time feedback from the equipment monitoring system, effectively addressing the uncertainties in multi-park operation and maintenance scenarios, and significantly improving work order processing efficiency, resource utilization, and the overall robustness and adaptability of the system.

[0015] In one or more embodiments of this application, the method for extracting feature vectors in step S200 includes: matching the corresponding skill requirement level from the equipment topology attributes based on the equipment type; matching the inventory status of the required parts from the parts topology attributes based on the fault type; and calculating the energy consumption impact coefficient of operation and maintenance work by combining the park-building-space and equipment energy consumption level.

[0016] By directly anchoring the extraction of feature vectors to the specific attributes of the digital twin topology, the real-time nature and accuracy of the feature data are ensured, providing a reliable data foundation for subsequent intelligent matching and decision-making, and avoiding decision-making errors caused by data lag.

[0017] In one or more embodiments of this application, the method for generating the initial dispatch plan in step S300 includes: using "equipment-parts-maintenance personnel" as matching nodes in the six-level topology, real-time monitoring of the parts warehouse inventory; if the required quantity of parts is greater than the real-time inventory of the parts warehouse, the corresponding equipment node of the part is removed, an abnormal signal of the maintenance work order is uploaded, and a procurement reminder is triggered in reverse; if the required quantity of parts is less than or equal to the real-time inventory of the parts warehouse, maintenance personnel with corresponding skill certifications are matched according to skill requirements and energy consumption sensitivity to form an initial pairing of equipment and maintenance personnel, thereby obtaining the initial dispatch plan.

[0018] By introducing spare parts inventory as a "hard constraint" for generating dispatch plans, invalid dispatches and path waste caused by spare parts shortages are fundamentally eliminated. At the same time, procurement reminders are automatically triggered, realizing efficient linkage between operation and maintenance and the supply chain, and improving the closed-loop automation level of the operation and maintenance process.

[0019] In one or more embodiments of this application, the method for generating the preliminary path planning for maintenance personnel in step S400 includes: constructing a directed graph using the three-level topological points of "park-building-space" in the six-level topology as path vertices and the internal roads of the park, underground parking lanes, and building elevators as edges, and writing real-time weights for each edge. The real-time weights are jointly calculated by the traffic flow density obtained and processed by the equipment monitoring system, the real-time occupancy rate of charging piles provided by the microgrid system, and the energy consumption prediction curve obtained by the energy consumption management system; establishing a first multi-objective optimization function based on the directed graph, the first multi-objective optimization function including at least two of the following: minimizing the skill matching difference, minimizing the accessory acquisition time, minimizing the overall path planning time, and minimizing the energy consumption impact coefficient; and obtaining the preliminary path planning by solving the first multi-objective optimization function.

[0020] By constructing a directed graph that integrates real-time data from multiple sources (vehicle flow, charging piles, and energy consumption) and performing multi-objective optimization, the generated initial path is no longer a static shortest path, but a high-quality solution that comprehensively considers efficiency, energy consumption, and feasibility, laying a solid foundation for subsequent dynamic optimization.

[0021] In one or more embodiments of this application, step S500, which dynamically adjusts the initial dispatching scheme based on the parameters fed back by the equipment monitoring system in real time to obtain an optimized dispatching scheme, includes: evaluating the equipment status and maintenance personnel status in real time according to the equipment monitoring system. The equipment status includes at least one of new equipment failure and equipment failure priority; the maintenance personnel status includes at least one of location change, work progress, and load status. When the maintenance required for the new equipment failure is consistent with the maintenance required at present, the maintenance is marked as "multi-device linkage", and other work order paths currently being executed are searched first. If there is a maintenance personnel about to enter the space and has new fault skills, their next node is merged into this node to complete batch maintenance. An optimized dispatching scheme is generated based on the equipment status and maintenance personnel status.

[0022] By identifying opportunities for "multi-device linkage" in real time and intelligently merging maintenance tasks, the repetitive movement of maintenance personnel and repeated start-up and shutdown of equipment can be effectively reduced, enabling batch maintenance, thereby significantly reducing maintenance costs and time, and minimizing interference with equipment operation.

[0023] In one or more embodiments of this application, step S600, which adjusts the preliminary path planning based on the optimized dispatch scheme to obtain the optimized path, includes: when a new node marked as "multi-device linkage" is inserted, immediately taking the current location of the maintenance personnel as the starting point, calling the "space-device-parts" sub-topology in the sixth-level topology, recalculating the minimum cost path, and adding a parts-in-the-way coefficient to the multi-objective optimization function to obtain a second multi-objective optimization function; if the next parts warehouse is located within the preset sector connecting the starting point and the new node, then forced to pick up the parts along the way; solving the second multi-objective optimization function to obtain the optimized path.

[0024] By creatively introducing a "parts-on-the-way coefficient" during path replanning, and by forcing the collection of parts along the same route in the optimization function, multiple task nodes are cleverly linked into an efficient path, further reducing unnecessary detours and improving the overall efficiency of a single trip.

[0025] In one or more embodiments of this application, step S600, which adjusts the preliminary path planning based on the optimized dispatch scheme to obtain the optimized path, further includes: when the load rate of maintenance personnel exceeds a certain preset value, and the insertion of a new node causes the path length to exceed the limit, a load balancing sub-loop is activated; the load balancing sub-loop means searching for other maintenance personnel in the sixth-level topology who are idle and hold the same skill tag, and if the straight-line distance between their current position and the new node is less than a certain proportion of the distance between the original maintenance personnel and the new node, then the new node and subsequent adjacent nodes are migrated to the idle maintenance personnel who hold the same skill tag, and a dual path is regenerated; the dual path is used as the optimized path.

[0026] By designing a "load balancing sub-loop" mechanism, when a single operations and maintenance (O&M) worker is overloaded, tasks can be intelligently transferred to more suitable, idle personnel. This avoids overloading of individual O&M workers, achieves dynamic balance of team workload, and ensures the timeliness and quality of O&M services.

[0027] Another aspect of this application provides a path planning system for equipment maintenance work orders across multiple campuses, comprising:

[0028] The digital twin module is used to construct and maintain a six-level topology of "park-building-space-equipment-parts-maintenance personnel" and write dynamic attribute vectors in real time; the feature vector engine module is used to receive maintenance work orders, extract skill requirements, parts requirements, and energy consumption sensitivity based on equipment type, fault type, and energy consumption levels of the park-building-space, forming a computable feature vector; the initial dispatch and inventory hard constraint module is used to eliminate missing parts nodes with real-time parts warehouse inventory as a hard constraint and trigger procurement reminders in reverse, while generating an initial dispatch plan based on skill certification and energy consumption sensitivity; the multi-objective path planning module is used to construct a directed graph with internal roads, underground parking lanes, and building elevators as edges, and to plan the path based on traffic flow. The edge weights are jointly calculated based on density, charging pile occupancy rate, and energy consumption prediction curves. The initial path planning is output by solving the first multi-objective optimization function. The real-time linkage optimization module is used to dynamically mark "multi-device linkage" nodes and insert them into the current path based on new fault signals from the equipment monitoring system, personnel location, and load status during the operation and maintenance personnel's execution. At the same time, the second multi-objective optimization function is called to recalculate the minimum cost path to obtain the optimized path. The path distribution and verification module is used to distribute the final optimized path to the operation and maintenance personnel and write the path results, parts collection records, and customer feedback information into the verification module. The verification module is used to verify whether the integration of inventory, operation and maintenance personnel, path, and energy consumption is optimal.

[0029] Through modular design, the system visualizes the key steps in the method into functionally defined and collaborative software and hardware modules. In particular, the "real-time linkage optimization module" enables online collaborative optimization of order dispatch and routes, and the "verification module" evaluates the overall optimization effect, forming a complete intelligent system integrating decision-making, execution, and verification.

[0030] In one or more embodiments of this application, the verification module includes a blockchain evidence storage module. The blockchain evidence storage module adopts a one-time write + multiple verification mode. The on-chain data structure includes path hash, parts inventory change, energy consumption reduction estimate and customer receipt signature. The verification interface is opened to the energy diagnosis subsystem of the energy consumption management module. It completes the operation and maintenance-energy saving causal verification within a certain period of time without accessing the original work order. If the verification result shows that the energy consumption reduction is lower than the preset threshold, the system triggers a "path rollback" signal, so that the next round of planning will prioritize the use of historical high energy-saving path templates, thereby achieving a reliable closed-loop control of energy-saving effect.

[0031] By leveraging the immutability of blockchain technology, a reliable mechanism for storing and verifying the energy-saving effects of operation and maintenance paths is provided. This "path rollback" mechanism enables the system to learn from historical experience, continuously selecting and solidifying high-energy-saving paths, thus achieving reliable closed-loop control and continuous optimization of energy-saving effects.

[0032] In one or more embodiments of this application, the path planning system further includes: an equipment management module, an equipment monitoring module, a microgrid module, an energy consumption management module, an operation and maintenance service module, and a customer management module; wherein, the multi-objective path planning module communicates in real time with the equipment management module, the equipment monitoring module, and the microgrid module, and the verification module communicates in real time with the energy consumption management module, the operation and maintenance service module, and the customer management module.

[0033] By clarifying the real-time communication relationships between various functional modules within the system, seamless data flow between multiple systems is ensured, thereby truly breaking down traditional information silos and realizing end-to-end data-driven decision-making from equipment monitoring to customer feedback, fully leveraging the synergistic effect of system integration.

[0034] The technical solution of this application will be further described below through specific implementation methods. Attached Figure Description

[0035] Figure 1 A flowchart illustrating a method for dispatching equipment maintenance work orders across multiple campuses, provided as an embodiment of this application;

[0036] Figure 2 This is a structural diagram of a multi-campus equipment maintenance work order path planning system provided in one embodiment of this application;

[0037] Figure 3 This is a schematic diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0038] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0039] To make the purpose, technical solution, and advantages of this application clearer, the following will be described in conjunction with the appendix. Figure 1-3 The following is an explanation using specific examples.

[0040] Please refer to Figure 1 , Figure 1 A flowchart illustrating a method for dispatching equipment maintenance work orders across multiple campuses, as provided in an embodiment of this application, includes the following steps:

[0041] S100: Construct a multi-campus digital twin model. The digital twin model is configured with a six-level topology of "campus-building-space-equipment-parts-maintenance personnel" and dynamic attribute vectors are written for each level of the topology.

[0042] S200: Receive maintenance work orders, extract key features based on the maintenance work orders, and break down the maintenance work orders into feature vectors including skill requirements, parts requirements, and energy consumption sensitivity based on the key features;

[0043] S300: Generates an initial dispatch scheme based on a six-level topology and feature vectors;

[0044] S400: Based on the initial dispatch plan, perform preliminary route planning for maintenance personnel;

[0045] S500: During the execution of the initial route plan by the maintenance personnel, the initial dispatch plan is dynamically adjusted based on the parameters fed back in real time by the equipment monitoring system to obtain an optimized dispatch plan;

[0046] S600: The optimized path is obtained by adjusting the initial path planning based on the optimized dispatching scheme;

[0047] S700: Maintenance personnel perform maintenance on the equipment based on optimized routes and optimized dispatching schemes.

[0048] In step S100, a multi-campus digital twin model is constructed. The digital twin model is configured with a six-level topology of "campus-building-space-equipment-parts-maintenance personnel" and a dynamic attribute vector is written for each level of topology.

[0049] As can be understood, a digital twin model refers to a high-fidelity virtual model created through digital means for a park, building, space, equipment, components, and maintenance personnel. This model not only reflects its geometric appearance but, more importantly, maps its real-time status (such as location, energy consumption, and inventory), behavior (such as operating modes), and rules (such as topological relationships), enabling simulation, analysis, and control. Multiple parks refer to geographically dispersed park entities that may be administratively independently managed.

[0050] As the top-level structure, the park-level topology undertakes macro-management functions. In actual deployment, each park node contains a unique identifier to ensure the consistency of cross-system data exchange; geographic boundary information is defined through high-precision GIS coordinates, supporting spatial analysis and cross-park path planning; park type identifiers distinguish different types such as production, R&D, and office, facilitating the implementation of differentiated operation and maintenance strategies. Furthermore, the park-level topology maintains connections with other parks, including physical connections (such as dedicated roads) and logical connections (such as management affiliation), providing a foundation for cross-park resource scheduling.

[0051] Building-level topology constructs a second layer of structure within the park. Each building node not only records basic building information (such as building area, number of floors, and year of construction), but more importantly, establishes a precise spatial relationship with other nodes in the park. Through BIM model integration, the building topology can provide detailed structural information, including load-bearing structures, pipeline layouts, and equipment room locations. This information is invaluable for scenarios such as operation and maintenance route planning, emergency evacuation, and replacement of large equipment. Building nodes also maintain energy supply relationships, recording critical infrastructure information such as the locations of major power access points and main water supply pipes.

[0052] Spatial topology further refines the management granularity, dividing buildings into spatial units with specific functions. Each spatial node clearly defines its functional type (such as office area, production workshop, equipment room, etc.) and records detailed spatial attributes, including space area, floor height, load-bearing capacity, and environmental requirements. The spatial topology also establishes complete adjacency and access relationships, recording connecting elements such as doors, corridors, and passageways, providing a foundation for indoor path planning. Crucially, spatial nodes maintain their containment relationships with equipment nodes, ensuring that each piece of equipment can be accurately associated with its actual spatial location.

[0053] Device-level topology is the core business node of the entire digital twin model. Each device node establishes a complete device profile, including static attributes such as device model, specifications, and technical documentation. More importantly, device nodes update dynamic operational data in real time through IoT interfaces: operating status (running, down, faulty), performance parameters (temperature, pressure, vibration, etc.), energy consumption data, maintenance history, etc. Device nodes also maintain complex relationships, including functional dependencies (e.g., a failure of device A will cause device B to stop), spatial relationships, and maintenance responsibility relationships. The explicit definition of these relationships provides a guarantee for the analysis of the impact range of device failures.

[0054] The component-level topology enables refined material management. Each component node records complete information about the component: model, specifications, applicable equipment, inventory quantity, and inventory location. Through RFID or QR code technology, the system tracks the flow of components in real time, including full lifecycle information such as warehousing, outbound, transfer, and inventory checks. Component nodes also maintain relationships with equipment nodes, clearly defining the applicable equipment range for each component. This relationship provides a basis for component availability checks during intelligent order dispatch. Furthermore, the component topology records supply chain data such as supplier information, procurement cycles, and minimum inventory levels, supporting inventory alerts and automatic replenishment.

[0055] The operations and maintenance personnel-level topology enables digital management of human resources. Each personnel node establishes a complete skill profile, including professional skill certifications, equipment operation qualifications, and historical work order completion status. Through mobile terminals and positioning systems, information such as personnel location status, workload, and current tasks is updated in real time. Personnel nodes also maintain organizational relationships, including their team affiliation, management reporting relationships, and collaboration relationships. This information provides a basis for decision-making regarding skill matching, location optimization, and load balancing during work order allocation.

[0056] This application innovatively introduces a six-level topology structure: "Park-Building-Space-Equipment-Parts-Maintenance Personnel". This six-level topology constructs a holographic digital mapping from the macro-level park to the micro-level maintenance elements, achieving multi-granularity precise positioning and dynamic association of maintenance resources. It breaks down data barriers between traditional independent management systems, providing a unified spatial-resource-status data foundation for intelligent work order dispatch and path planning, thereby significantly improving the collaborative efficiency and resource utilization of cross-park maintenance operations.

[0057] Dynamic attribute vectors refer to a set of state parameters that change over time. Their technical characteristics include: 1) multi-dimensional, covering multiple aspects such as operating status, performance indicators, and environmental conditions; 2) temporal sequence, retaining historical change trajectories and supporting trend analysis; 3) real-time, achieving near real-time state updates through sensor networks; and 4) correlation, with mutual influence relationships between different attributes.

[0058] Dynamic attribute vectors are the data foundation for intelligent decision-making in a system, and their quality and timeliness directly affect the effectiveness of operational decisions. The implementation of dynamic attribute vectors relies on a multi-layered sensor network. At the physical layer, various sensors are deployed, including environmental sensors (temperature, humidity, air quality, illuminance), equipment sensors (vibration, temperature, pressure, current), and location sensors (GPS, UWB, RFID). These sensors collect and perform preliminary data processing through an IoT gateway, forming a standardized data stream.

[0059] In step S200, an operation and maintenance work order is received, and key features are extracted based on the operation and maintenance work order. Based on the key features, the operation and maintenance work order is split into feature vectors including skill requirements, spare parts requirements, and energy consumption sensitivity.

[0060] In one or more embodiments of this application, the method for extracting feature vectors in step S200 includes: matching the corresponding skill requirement level from the equipment topology attributes based on the equipment type; matching the inventory status of the required parts from the parts topology attributes based on the fault type; and calculating the energy consumption impact coefficient of operation and maintenance work by combining the park-building-space and equipment energy consumption level.

[0061] Equipment topology attributes refer to various parameters defined in the equipment-level nodes of the digital twin model, including static and dynamic information such as equipment model, specifications, operating status, and maintenance records. Parts topology attributes refer to inventory information recorded in the parts-level nodes of the digital twin model, including real-time data such as parts model, inventory quantity, storage location, and inventory status. The energy consumption impact coefficient is a comprehensive evaluation indicator that quantitatively analyzes the degree of impact of operation and maintenance tasks on system energy consumption, helping to optimize task scheduling and resource allocation.

[0062] Specifically, the extraction process of skill requirement features relies heavily on the attribute data of the device-level topology in the digital twin model. When the system receives a maintenance work order, it first identifies the device type from the work order description, and then queries the device topology for the maintenance skill requirements of that type of device. The specific extraction process is as follows:

[0063] The system analyzes static attributes such as equipment model and specifications, combined with dynamic attributes such as equipment operating status and maintenance history, to comprehensively determine the skill level required to complete the work order. For example, when receiving a "centrifugal pump bearing replacement" work order, the system retrieves basic information such as the pump's model and power from the equipment topology, and simultaneously queries the equipment's maintenance records. If the equipment operates under high pressure and high temperature conditions, and historical maintenance records indicate that special tools and processes are required, the system will set the skill requirement level to high.

[0064] During the skills matching process, the system also considers the real-time operating status of the equipment. For example, for equipment running on a critical production line, even if the maintenance work itself is not technically difficult, the system will appropriately increase the skill requirement level to ensure that experienced maintenance personnel are dispatched due to the significant impact of downtime.

[0065] The extraction of spare parts demand characteristics is based on the real-time correlation between fault type and spare parts topology. The system first analyzes the fault description in the work order to determine the types of spare parts that may be needed, and then immediately queries the inventory status in the spare parts topology. The specific implementation process is as follows:

[0066] When the work order description mentions "seal leakage," the system automatically links it to the demand for accessories such as sealing rings and gaskets. The system checks the inventory quantity and storage location of these accessories in the accessory topology in real time. For example, if there are 5 sets of a certain model of mechanical seal in stock, stored in warehouse A area 3 of building B in the park, this information will be updated to the feature vector in real time.

[0067] As a preferred approach, the system also establishes a spare parts substitution rule base. When the inventory of major spare parts is insufficient, it will automatically search for alternative spare parts models. For example, when a certain specification of bearing is out of stock, the system will look for alternative models with similar dimensions and the same load rating, and incorporate this information into the spare parts demand characteristics.

[0068] The assessment of energy consumption sensitivity characteristics is a multi-level calculation process. This application's embodiment innovatively integrates energy consumption characteristics at four levels: park, building, space, and equipment. The specific calculation process is as follows:

[0069] First, the system acquires overall energy consumption data from the campus-level topology, including current electricity prices and grid load. Second, it acquires building energy consumption characteristics from the building-level topology, such as air conditioning system operation modes and lighting system schedules. Then, it acquires environmental control requirements from the space-level topology, such as temperature and humidity control ranges and cleanliness levels. Finally, it combines data such as equipment power and operating time from the equipment-level topology to comprehensively calculate the energy consumption impact of maintenance operations. For example, during peak electricity consumption hours in the summer afternoon, maintenance work orders for a data center's air conditioning system will be assigned a high energy sensitivity. This is because maintenance work at this time may cause the computer room temperature to rise, thereby increasing the load on the cooling system and resulting in significant energy costs during peak electricity price periods. Conversely, the same maintenance work performed at night will have a significantly lower energy sensitivity.

[0070] Specifically, energy consumption sensitivity characteristics can be represented by energy consumption sensitivity. The specific expression for energy consumption sensitivity is:

[0071] Es = α·E_campus + β·E_building + γ·E_space + δ·E_device

[0072] Where Es represents energy consumption sensitivity;

[0073] E_campus represents the energy consumption characteristics of the campus. E_campus = f1(current electricity price, grid load factor, microgrid status) × campus energy consumption benchmark, where α is the weighting coefficient of the campus energy consumption characteristics.

[0074] E_building represents the building's energy consumption characteristics, E_building = f2(air conditioning operation mode, lighting schedule, building energy density) × building energy consumption baseline; β is the building energy consumption characteristic weighting coefficient;

[0075] E_space represents the space energy consumption characteristics, E_space = f3(temperature and humidity control range, cleanliness level, space function type) × space energy consumption benchmark; γ is the weighting coefficient of space energy consumption characteristics;

[0076] E_device represents the energy consumption characteristics of the device. E_device = f4(rated power of the device, estimated running time, energy efficiency level of the device) × energy consumption baseline of the device; δ is the weighting coefficient of the energy consumption characteristics of the device.

[0077] The park's energy consumption benchmark is based on the park's functional positioning, industry type, and overall energy consumption scale, reflecting the park's basic energy consumption level. This benchmark value embodies the basic energy consumption characteristics of the park as a whole energy-consuming unit and is directly related to the park's industry nature, building area, and energy-consuming equipment density. By analyzing the park's total energy consumption data over the past 12 months, combined with park type classification (such as industrial parks, science and technology parks, business parks, etc.), a normalized calculation is performed using a unit area energy consumption index. The specific formula is: Park Energy Consumption Benchmark = (Park's Annual Total Electricity Consumption / Park's Total Area) / Standard Energy Consumption per Unit Area for Similar Parks. For example, data center parks: 1.2, high-end manufacturing parks: 1.0, science and technology R&D parks: 0.8, general office parks: 0.5, warehousing and logistics parks: 0.4.

[0078] The building energy consumption baseline reflects the energy consumption characteristics of the building itself, including the performance of the building envelope, equipment system configuration, and usage patterns. This baseline value comprehensively reflects the combined impact of the building's thermal performance, equipment system efficiency, and operational management level.

[0079] Based on the energy consumption data of the building energy management system, combined with parameters such as building age, window-to-wall ratio, external wall insulation performance, and equipment system configuration, for example, a 24-hour data center consumes 1.2, a tertiary hospital inpatient building consumes 1.0, a Grade A office building consumes 0.8, a general office building consumes 0.6, and a light industrial factory consumes 0.5.

[0080] A spatial energy consumption baseline characterizes the environmental control requirements and energy intensity of a specific functional space, depending on the space's functional positioning, environmental control standards, and equipment configuration density. This baseline value reflects the inherent energy consumption requirements of the space's function. It is determined based on environmental control parameters (temperature and humidity range, cleanliness level, fresh air volume requirements, etc.) in the space design specifications, combined with the power density and operating time requirements of the equipment within the space, through spatial function classification and energy demand analysis. For example, a baseline value of 1.1 for a biological laboratory, 1.0 for a data center server room, 0.9 for a cleanroom, 0.5 for a conference room, 0.4 for a general office, and 0.2 for a corridor area.

[0081] Equipment energy consumption benchmarks reflect the equipment's energy efficiency characteristics and operating features, including rated power, operating time mode, and energy efficiency rating. This benchmark value embodies the basic energy consumption level of the equipment under normal operating conditions. It is determined through standardized evaluation based on equipment energy efficiency labels, manufacturer-provided technical parameter tables, equipment operation logs, equipment type, service life, and operating conditions, according to equipment energy efficiency grading standards. For example, large chillers have a benchmark of 1.2, server racks 1.0, precision air conditioners 0.9, elevator systems 0.7, LED lighting systems 0.3, and office computers 0.2.

[0082] α + β + γ + δ = 1, wherein, in a preferred embodiment, α:β:γ:δ = 3:3:2:2.

[0083] Energy consumption characteristics at each level are normalized by standardization mapping to the [0,1] interval.

[0084] When equipment maintenance is required in a data center in a certain park, the system will mark it as a high-energy-consumption sensitive operation because the data center has strict requirements for temperature and humidity control (high E_space value), high demand for air conditioning operation (high E_building value), high equipment power density (high E_device value), and may be in peak electricity consumption period (high E_campus value). It is recommended to schedule the operation during the night when the electricity price is low or during the period when the park's photovoltaic power generation is sufficient.

[0085] When maintenance is required on the office building's routine lighting system, the space requirements are relatively relaxed (low E_space value), the lighting equipment power is relatively small (medium E_device value), and if the work is carried out during normal power hours (medium E_campus value), the energy consumption sensitivity is low, and the system will generate a routine optimized path.

[0086] The system also considers the impact of job duration on energy consumption. For example, equipment maintenance work that is expected to take 4 hours will be assigned a higher energy sensitivity level than a routine inspection that only takes 30 minutes, because prolonged equipment downtime or degraded operation may result in greater energy waste.

[0087] Using the aforementioned feature extraction methods, the system can transform abstract maintenance work orders into concrete, quantifiable feature vectors. Skill requirement features ensure precise matching between maintenance tasks and personnel capabilities; spare parts requirement features guarantee a reliable supply of maintenance resources; and energy consumption sensitivity features support energy efficiency optimization during the maintenance process. These three feature vectors provide accurate decision-making basis for subsequent intelligent work order dispatching and route planning, significantly improving the intelligence level and execution efficiency of maintenance management.

[0088] In step S300, an initial dispatch plan is generated based on the six-level topology and feature vectors.

[0089] In one or more embodiments of this application, the method for generating the initial dispatch plan in step S300 includes: using "equipment-parts-maintenance personnel" as matching nodes in the six-level topology, real-time monitoring of the parts warehouse inventory; if the required quantity of parts is greater than the real-time inventory of the parts warehouse, the corresponding equipment node of the part is removed, an abnormal signal of the maintenance work order is uploaded, and a procurement reminder is triggered in reverse; if the required quantity of parts is less than or equal to the real-time inventory of the parts warehouse, maintenance personnel with corresponding skill certifications are matched according to skill requirements and energy consumption sensitivity to form an initial pairing of equipment and maintenance personnel, thereby obtaining the initial dispatch plan.

[0090] Specifically, the initial order dispatch plan is generated based on the hard constraint of parts availability. The system monitors the inventory status in the parts topology in real time and accurately compares the required parts quantity with the real-time inventory. The specific implementation process is as follows:

[0091] The system extracts parts demand information from the feature vector, including the type, specifications, and quantity of the required parts. It then immediately queries the real-time inventory data of the corresponding warehouse location in the parts topology. This process is completed through the following steps:

[0092] Parts requirement analysis: Analyze the parts requirement dimension in the feature vector to extract the specific parts codes and required quantities;

[0093] Inventory status query: Real-time availability of inventory for corresponding codes can be obtained through the topology attributes of the parts;

[0094] Availability assessment: Compare the required quantity with real-time inventory, and perform a strict numerical comparison;

[0095] For example, when a work order is received to "replace 3 sets of B01 model bearings", the system will: identify the need for 3 sets of B01 bearings from the parts demand characteristics, query the real-time inventory quantity of B01 in the parts topology, and if the inventory shows only 2 sets, it will be determined that the inventory is insufficient. The system will automatically remove the equipment node to avoid invalid order dispatch.

[0096] In the event of insufficient inventory, the system will immediately upload an anomaly signal to the maintenance work order and trigger the procurement process. The anomaly signal includes detailed out-of-stock information: out-of-stock part code, required quantity, current inventory, and suggested procurement quantity. A procurement reminder will be automatically sent to the procurement system, including an urgency assessment and expected delivery time.

[0097] After screening through spare parts inventory, the system enters the maintenance personnel matching phase. This phase comprehensively considers skill requirements and energy consumption sensitivity characteristics to achieve optimal personnel allocation. The specific matching process is as follows:

[0098] Skills requirement matching: The system obtains the skill requirement level from the feature vector and then searches for personnel with matching skills certifications within the operations and maintenance personnel topology. The matching process considers not only whether the corresponding skills are possessed but also the degree of fit between the skill level and the requirement level. For example, for a work order requiring "high-voltage electrician certificate + equipment-specific certification," the system will: screen all certified personnel, rank them based on factors such as historical work order completion quality and professional skills assessment scores, and prioritize the operations and maintenance personnel with the highest skill matching degree.

[0099] Building upon skill matching, the system further optimizes its performance by incorporating energy consumption sensitivity characteristics. By analyzing factors such as the maintenance personnel's current location, estimated job duration, and energy consumption impact during the job's duration, the system selects the dispatch plan with the lowest overall energy consumption. For example, during peak summer electricity consumption periods, a central air conditioning unit requires maintenance: the system identifies this job as having high energy consumption sensitivity and prioritizes personnel who are closer and can complete the task quickly, avoiding scheduling time-consuming and complex tasks during peak electricity price periods.

[0100] The system uses a weighted allocation model to balance two dimensions: skill matching and energy consumption sensitivity. For critical equipment or work orders with high energy consumption sensitivity, the weight of energy consumption sensitivity is appropriately increased; for general maintenance operations, more emphasis is placed on the accuracy of skill matching.

[0101] After completing the above screening and matching, the system generates the final initial equipment-maintenance personnel pairing. This process ensures that each work order is assigned to the most suitable maintenance personnel, while meeting the requirements of parts availability and energy efficiency optimization. The pairing rules include: one-to-one matching: ensuring each work order has a clearly defined responsible person; load balancing: considering the current workload of maintenance personnel; geographical location optimization: prioritizing personnel with closer proximity; skill expertise utilization: fully leveraging the professional expertise of personnel. For example: A park receives three work orders simultaneously: Work Order A: Replacement of water pump mechanical seal (requires advanced mechanic skills); Work Order B: Inspection of power distribution cabinet (requires intermediate electrician skills); Work Order C: Cleaning of air conditioning system (requires intermediate HVAC skills). The system will: first check the parts inventory for each work order, then match the appropriate personnel according to skill requirements, and optimize the allocation based on the personnel's current location and energy consumption sensitivity, generating three independent equipment-personnel pairing schemes.

[0102] By employing a screening mechanism based on hard constraints of spare parts inventory, the system effectively avoids invalid order dispatches caused by spare parts shortages, thus improving the initial order dispatch success rate. Intelligent matching, combining skill requirements and energy consumption sensitivity, ensures optimal alignment between maintenance tasks and personnel capabilities, while optimizing energy usage efficiency. This multi-dimensional approach to order dispatch generation significantly improves the utilization efficiency and job quality of maintenance resources, laying a solid foundation for subsequent path planning and dynamic optimization.

[0103] In step S400, preliminary route planning for maintenance personnel is carried out according to the initial dispatch plan.

[0104] In one or more embodiments of this application, the method for generating the preliminary path planning for maintenance personnel in step S400 includes: constructing a directed graph using the three-level topological points of "park-building-space" in the six-level topology as path vertices and the internal roads of the park, underground parking lanes, and building elevators as edges, and writing real-time weights for each edge. The real-time weights are jointly calculated by the traffic flow density obtained and processed by the equipment monitoring system, the real-time occupancy rate of charging piles provided by the microgrid system, and the energy consumption prediction curve obtained by the energy consumption management system; establishing a first multi-objective optimization function based on the directed graph, the first multi-objective optimization function including at least two of the following: minimizing the skill matching difference, minimizing the accessory acquisition time, minimizing the overall path planning time, and minimizing the energy consumption impact coefficient; and obtaining the preliminary path planning by solving the first multi-objective optimization function.

[0105] It can be understood that a directed graph is a graph structure composed of vertices and directed edges, used to accurately describe the topological relationships and directional constraints of a path network; the first multi-objective optimization function refers to the optimization model used in the initial path planning stage, which seeks the comprehensive optimal path solution by balancing multiple conflicting objectives.

[0106] The system constructs a directed graph structure for the path network based on a three-level topology of "campus-building-space". The specific construction process is as follows:

[0107] Path vertices are derived from spatial location nodes in the topology, including key coordinate points such as park entrances / exits, building lobbies, and specific spatial locations. For example, the main entrance of Park A, the first-floor lobby of Building B, and the 201 computer room of Building C are all defined as path vertices. Path edges are established based on actual travel paths, including: internal park roads: external passages connecting different buildings; underground parking garage driveways: driving passages in the underground parking garage; building elevators: vertical passages connecting different floors; fire exits: backup paths in emergency situations. Each edge contains direction information, accurately recording restrictions such as one-way and two-way traffic. For example, the road from Building A to Building B may be two-way, but a certain underground parking garage exit may be one-way.

[0108] The system calculates a real-time weight for each edge, which is a comprehensive cost indicator. The real-time weight is jointly calculated from the traffic density obtained and processed by the equipment monitoring system, the real-time occupancy rate of charging piles provided by the microgrid system, and the energy consumption prediction curve obtained by the energy management system. Specifically, the traffic density dimension is obtained from real-time traffic data acquired by the equipment monitoring system, including the number of vehicles identified by cameras, the traffic speed detected by geomagnetic sensors, and historical traffic pattern data for the same period. For example, if the traffic density of a certain road in the park is 85% during the morning rush hour, the system will correspondingly increase the traffic weight of that road segment, guiding users to choose alternative routes. The charging pile occupancy rate dimension is obtained from the real-time status of charging piles acquired by the microgrid system, including the number of available charging piles, estimated waiting time, and charging power limits. For example, if maintenance personnel need to charge their vehicles, the system will prioritize routes in areas with lower charging pile occupancy rates. The energy consumption prediction dimension is obtained from the energy management system, including prediction data such as time-of-use electricity price curves, equipment operating energy consumption patterns, and the influence coefficient of ambient temperature. For example, during peak electricity price periods, the system assigns higher weights to routes passing through high-energy-consuming areas, guiding users to choose more energy-efficient routes. This application innovatively introduces real-time weights into the construction of the directed graph network, enabling the system to accurately reflect the actual traffic conditions of routes within the park and quickly adapt to environmental changes, ensuring the accuracy and practicality of route planning.

[0109] The first multi-objective optimization function established by the system includes:

[0110] Minimize skill matching discrepancies: Based on the skill matching results in the initial dispatch plan, evaluate the impact of path selection on skill utilization efficiency. For example, prioritize paths that fully leverage the expertise of operations and maintenance personnel, even if those paths are slightly longer.

[0111] Minimize parts retrieval time: Consider the location distribution of parts warehouses and optimize parts retrieval paths. For example, if a work order requires parts to be retrieved from multiple warehouses, the system will optimize the access order and reduce the total waiting time.

[0112] Minimize the total path time: comprehensively estimate the travel time of each path segment, including: road travel time, elevator waiting time, security check time, and equipment operation preparation time.

[0113] Minimize the energy consumption impact factor: Evaluate the energy consumption impact of a route by incorporating energy consumption sensitivity characteristics. For example, avoid scheduling routes through high-energy-consuming areas during peak electricity consumption periods.

[0114] The application of the first multi-objective optimization function ensures that path planning achieves a balance across multiple dimensions, including efficiency, cost, and resource utilization. This refined path planning method significantly improves the mobility efficiency of operations and maintenance personnel, reduces operating costs, and simultaneously guarantees the timeliness and quality of operations and maintenance services.

[0115] In step S500, while the maintenance personnel are executing the initial path plan, the initial dispatch plan is dynamically adjusted based on the parameters fed back by the equipment monitoring system to obtain an optimized dispatch plan.

[0116] In one or more embodiments of this application, step S500, which dynamically adjusts the initial dispatching scheme based on the parameters fed back by the equipment monitoring system in real time to obtain an optimized dispatching scheme, includes: evaluating the equipment status and maintenance personnel status in real time according to the equipment monitoring system. The equipment status includes at least one of new equipment failure and equipment failure priority; the maintenance personnel status includes at least one of location change, work progress, and load status. When the maintenance required for the new equipment failure is consistent with the maintenance required at present, the maintenance is marked as "multi-device linkage", and other work order paths currently being executed are searched first. If there is a maintenance personnel about to enter the space and has new fault skills, their next node is merged into this node to complete batch maintenance. An optimized dispatching scheme is generated based on the equipment status and maintenance personnel status.

[0117] This step is implemented based on a complete real-time data acquisition, intelligent analysis, and dynamic decision-making system. The system continuously monitors key operating parameters, including vibration frequency, temperature changes, and pressure fluctuations, through a sensor network deployed at the device level. Simultaneously, it tracks the geographical location, task execution progress, and workload status in real time through smart terminals worn by maintenance personnel. The specific operation process is as follows: When the system detects a new fault signal through the device monitoring network, it first extracts fault features and assesses their severity, comparing them with the existing work order database from multiple dimensions. The system focuses on examining the correlation between the new fault and ongoing work orders in terms of skill requirements, spatial location, and time window, calculating the feasibility of task merging using a pre-set algorithm model. When a suitable linkage opportunity is identified, the system initiates a "multi-device linkage" processing flow, searching for maintenance personnel currently performing similar tasks, analyzing the adjustment space of their task sequences, and replanning the task execution order based on optimization algorithms. Throughout the process, the adjustment plan is pushed to relevant personnel through a real-time communication mechanism, and supporting arrangements such as parts allocation and tool preparation are updated simultaneously to ensure the smooth implementation of the optimization plan.

[0118] The system operates based on dynamic resource optimization theory and intelligent decision-making models. It establishes a real-time mapping between equipment status and operational resources, employing combinatorial optimization algorithms from operations research to find the optimal task allocation scheme under multiple constraints. Specifically, the system uses graph theory to analyze the topological relationships between devices, calculating spatial distance and functional dependencies; it assesses the workload and task processing capabilities of maintenance personnel based on queuing theory; and it uses heuristic algorithms to solve the task sequence optimization problem. In multi-device linkage identification, the system uses feature matching algorithms to find faulty devices with similar skill requirements, utilizes spatial clustering analysis to determine the proximity relationships between devices, and combines time series prediction to evaluate the temporal feasibility of task execution. The process of generating an optimized dispatch scheme is essentially a multi-objective decision problem. The system needs to seek a balance among multiple objectives such as response speed, resource utilization, and service cost, obtaining the optimal solution through real-time calculation.

[0119] To illustrate the actual operation of this process, consider a specific case: Two chiller units in a central air conditioning system in a certain industrial park simultaneously triggered fault alarms. Real-time monitoring data revealed that the compressor vibration of Unit 1 exceeded the standard, and the oil temperature of Unit 2 was abnormally high. Both faults required maintenance personnel with "large refrigeration equipment repair" skills to handle. The system detected that Engineer Zhang was nearby handling a fan coil unit fault and was about to complete his current task. Therefore, the system marked the two chiller unit faults as "multi-device linkage" and generated a new dispatch plan: After completing his current task, Engineer Zhang would first address the vibration problem of Unit 1, and then address the abnormal oil temperature of Unit 2. Simultaneously, the system notified the parts warehouse to prepare the corresponding seals and lubricants and optimized the tool allocation plan. Through this dynamic adjustment, tasks that would have previously required two separate dispatches were combined into a single continuous operation, reducing the travel time of maintenance personnel and improving work efficiency.

[0120] The significant technical benefits of this dynamic dispatch optimization method are reflected in improved operational efficiency and reduced costs. Through real-time monitoring and intelligent analysis, the system can promptly identify task merging opportunities, consolidating scattered maintenance tasks into batch processing, effectively reducing personnel travel time and repetitive preparation work. The multi-device linkage mechanism allows for the full utilization of professional skills, enabling the same maintenance personnel to handle multiple similar faults in a single operation, improving the efficiency of human resource utilization. The system's real-time adjustment capability ensures rapid response in the event of sudden failures, maintaining service continuity through intelligent redispatch. Statistics show that this dynamic optimization mechanism increases the effective working time of maintenance personnel, reduces travel distance, and significantly improves customer satisfaction due to the one-time resolution of problems.

[0121] In step S600, the preliminary route planning is adjusted based on the optimized dispatching scheme to obtain the optimized route.

[0122] In one or more embodiments of this application, step S600, which adjusts the preliminary path planning based on the optimized dispatch scheme to obtain the optimized path, includes: when a new node marked as "multi-device linkage" is inserted, immediately taking the current location of the maintenance personnel as the starting point, calling the "space-device-parts" sub-topology in the sixth-level topology, recalculating the minimum cost path, and adding a parts-in-the-way coefficient to the multi-objective optimization function to obtain a second multi-objective optimization function; if the next parts warehouse is located within the preset sector connecting the starting point and the new node, then forced to pick up the parts along the way; solving the second multi-objective optimization function to obtain the optimized path.

[0123] In one or more embodiments of this application, step S600, which adjusts the preliminary path planning based on the optimized dispatch scheme to obtain the optimized path, further includes: when the load rate of maintenance personnel exceeds a certain preset value, and the insertion of a new node causes the path length to exceed the limit, a load balancing sub-loop is activated; the load balancing sub-loop means searching for other maintenance personnel in the sixth-level topology who are idle and hold the same skill tag, and if the straight-line distance between their current position and the new node is less than a certain proportion of the distance between the original maintenance personnel and the new node, then the new node and subsequent adjacent nodes are migrated to the idle maintenance personnel who hold the same skill tag, and a dual path is regenerated; the dual path is used as the optimized path.

[0124] This step is implemented based on dynamic path replanning and intelligent resource scheduling. When the system identifies a new node in a "multi-device linkage" scenario that needs to be inserted into the existing task sequence, it immediately initiates a real-time path optimization mechanism. The system first obtains the current location of the maintenance personnel as the starting point for path planning, and then calls the "space-device-parts" sub-topology relationship network in the six-level topology. This sub-topology accurately describes the spatial location of the devices, the distribution of parts warehouses, and the connectivity between them. When recalculating the minimum cost path, the system innovatively adds a key indicator, the "parts route coefficient," to the original multi-objective optimization function, forming a second multi-objective optimization function. This coefficient is calculated using a spatial geometry algorithm: using the line connecting the current location of the maintenance personnel and the location of the new task node as the reference axis, a sector area with an angle range of ±30 degrees is established. The system detects the location of all parts warehouses within this sector. If the required parts are found to be located within this sector, the system will force the parts pickup point to be inserted into the path sequence, thereby achieving the optimization effect of "pickup along the route." In actual path calculation, the system comprehensively considers multiple dimensions such as path length, time cost, and ease of obtaining spare parts, and obtains the optimal path through iterative optimization.

[0125] The system operates based on spatial optimization theory and dynamic resource scheduling algorithms. By establishing a precise spatial topological network, the complex path planning problem is transformed into a shortest path problem in graph theory. The introduction of the accessory route coefficient essentially adds a consideration of resource acquisition efficiency to the traditional distance optimization, demonstrating the application of multi-objective decision theory in practical engineering. The spatial sector region determination method used by the system is based on the ray detection principle in computer graphics. It quickly determines location relevance by calculating the angular relationship between points and line segments, achieving high computational efficiency while maintaining accuracy. At the path optimization algorithm level, the system transforms the time cost of accessory retrieval into part of the path weight, using dynamic programming to solve for the optimal sequence that comprehensively considers movement time and resource acquisition time.

[0126] To illustrate the process, consider this concrete example: While an operations and maintenance (O&M) worker is troubleshooting an equipment malfunction in Building A, the system detects a new equipment malfunction in Building B requiring replacement of a specialized seal. Through spatial topology analysis, the system discovers that the C parts warehouse is located within the fan-shaped area of ​​the path from Building A to Building B, requiring only a 2-minute detour from the main path. The system immediately recalculates the path, inserting the C warehouse as a mandatory stop in the path sequence. Simultaneously, the system notifies the warehouse in advance to prepare the necessary seal, ensuring the O&M worker can retrieve it directly upon arrival. Although this optimization increases the total path length, the overall task time is reduced by avoiding a dedicated parts retrieval journey.

[0127] During the implementation of the load balancing mechanism, the system continuously monitors the workload rate of operations and maintenance (O&M) personnel. When it detects that an O&M personnel's load exceeds a preset threshold (e.g., 85%), and the insertion of a new task would cause the path length to exceed the limit, the system automatically starts a load balancing sub-loop. The core of this mechanism is to quickly search for O&M personnel with the same skill tags who are in an idle state within the six-level topology, and calculate the straight-line distance between each candidate personnel and the new task node using a spatial indexing algorithm. The system sets a key percentage value: when the distance between other O&M personnel and the new task node is less than a certain percentage value of the distance between the original personnel and the new node, task migration is triggered. This process not only migrates individual new task nodes but also analyzes task correlations, migrating subsequent nodes along the same path as a whole to ensure that the task sequence of the newly received personnel is also optimized. The system generates optimized path schemes for both the original O&M personnel and the newly received personnel, forming a collaborative "dual-path" architecture.

[0128] The R&D team selected a specific percentage for the load balancing mechanism based on extensive analysis and experimental verification of real-world operational data. Experimental data showed that while a percentage below 50% ensured closer proximity for newly assigned personnel, it could lead to increased coordination costs due to excessive task splitting. Conversely, a percentage above 70% might result in overly lenient migration conditions, impacting the original plan's execution. Comparative analysis of three months of operational data revealed that a percentage between 50% and 70% achieved a relatively optimal balance between task completion efficiency and resource utilization.

[0129] In this embodiment, an innovative "load balancing sub-loop" mechanism is designed to automatically redistribute new task packages to available personnel with closer geographical locations and matching skills when an excessive load is detected in a single maintenance worker. This effectively avoids response delays caused by excessive task concentration. At the same time, the dual-path collaborative optimization achieves real-time balance of team resources and improves overall maintenance efficiency.

[0130] The technical benefits of this route optimization method are evident in multiple dimensions. Through intelligent route optimization, the system integrates parts acquisition time into route planning, reducing dedicated material allocation trips and improving time utilization. The load balancing mechanism ensures the rational allocation of resources, preventing situations where individual personnel are overloaded while others are idle. Simultaneously, due to the improved rationality of route planning, the energy consumption of maintenance vehicles is also reduced, achieving a dual optimization of maintenance efficiency and cost control. The entire system, through dynamic route adjustment and resource scheduling, establishes an efficient, flexible, and adaptive maintenance service system.

[0131] The first and second multi-objective optimization functions in this application can be solved using existing algorithms such as evolutionary algorithms, mathematical programming, and metaheuristic algorithms. The embodiments in this application are not explained in detail.

[0132] See Figure 2In another embodiment of this application, a path planning system for equipment maintenance work orders across multiple parks is provided, comprising: a digital twin module 21, used to construct and maintain a six-level topology of "park-building-space-equipment-parts-maintenance personnel" and write dynamic attribute vectors in real time; a feature vector engine module 22, used to receive maintenance work orders and extract skill requirements, parts requirements, and energy consumption sensitivity based on equipment type, fault type, and energy consumption level of park-building-space, forming a computable feature vector; and an initial dispatch and inventory hard constraint module 23, used to eliminate missing parts nodes based on real-time inventory of parts warehouses as a hard constraint and trigger procurement reminders in reverse. Simultaneously, an initial dispatch plan is generated based on skill certification and energy consumption sensitivity; the multi-objective path planning module 24 is used to construct a directed graph with internal roads, underground parking lanes, and building elevators as edges, and calculate edge weights jointly with traffic density, charging pile occupancy rate, and energy consumption prediction curve, and output the preliminary path plan by solving the first multi-objective optimization function; the real-time linkage optimization module 25 is used to dynamically mark "multi-device linkage" nodes and insert them into the current path based on new fault signals from the equipment monitoring system, personnel location, and load status during the execution of operations and maintenance personnel, and simultaneously call the second multi-objective optimization function to recalculate the minimum cost path to obtain the optimized path; the path distribution and verification module 26 is used to distribute the final optimized path to operations and maintenance personnel, and write the path results, parts collection records, and customer feedback information into the verification module. The verification module is used to verify whether the integration of inventory-operations and maintenance personnel-path-energy consumption is optimal.

[0133] This system is implemented through a deep integration of digital twin technology and multi-objective optimization algorithms. At the park level, GPS positioning base stations and environmental monitoring stations are deployed to acquire spatial layout and environmental parameters in real time. At the building level, BIM models are integrated to record building structure information and energy supply relationships. At the spatial level, temperature and humidity sensors and access control systems track space usage status. At the equipment level, PLC controllers are connected to monitor equipment operating parameters. At the parts level, RFID technology is used to manage inventory status. At the maintenance personnel level, smart badges and mobile terminals are used to obtain real-time location and work status. After cleaning and standardization, this data is stored in a graph database and a time-series database, forming a complete multi-dimensional topological relationship network. The core principle of this module is to establish a precise mapping between physical entities and digital models, providing an accurate and real-time data foundation for subsequent intelligent decision-making, ensuring that the system can perform optimization calculations based on real-world conditions.

[0134] The feature vector engine module 22 employs natural language processing and machine learning technologies to transform unstructured maintenance work orders into structured feature vectors. Upon receiving a maintenance work order, the system first performs text semantic analysis to identify key entities and relationships. Then, it matches skill requirement levels based on the equipment knowledge graph. For example, for a "centrifugal pump bearing replacement" work order, the system queries the maintenance requirements for that pump model and determines the required skill level based on the equipment's operating status. Parts requirement features are generated through association rule mining; the system analyzes the correlation between fault types and parts usage in historical work orders and adjusts them based on real-time inventory status. Energy consumption sensitivity features are derived through multi-level calculations, comprehensively considering factors such as equipment power, operation time, and spatial environment requirements. The principle of this module lies in quantifying complex maintenance requirements into calculable standard indicators through feature engineering techniques, providing a unified input format for subsequent optimization decisions.

[0135] The initial order dispatch and inventory hard constraint module implements an intelligent filtering mechanism based on resource availability. This module first rigorously verifies spare parts requirements by querying inventory data in the spare parts topology via a real-time API interface. When it detects that 3 sets of a certain model of mechanical seal are needed but only 2 sets are in stock, the system automatically removes that equipment node to avoid invalid order dispatch and triggers the procurement process. After passing inventory verification, the module matches personnel based on skill requirements and energy consumption sensitivity, using a multi-objective optimization algorithm to select the maintenance personnel with the lowest overall cost while meeting skill requirements. The operating principle of this module is to integrate resource constraints into the decision-making front end, ensuring the feasibility of the order dispatch plan through hard condition filtering, and then improving the economic efficiency of the plan through soft optimization.

[0136] The multi-objective path planning module 24 constructs a dynamically weighted directed graph network. This module uses internal park roads, underground parking garage lanes, and building elevators as edges and spatial locations as vertices to establish a complete path network model. Edge weights are calculated in real-time using multi-source data: traffic density data comes from cameras and geomagnetic sensors, charging pile occupancy rate data comes from the microgrid system, and energy consumption prediction curves come from the energy management system. The module employs an improved genetic algorithm to solve the first multi-objective optimization function, balancing multiple objectives such as path length, time cost, and energy consumption impact. For example, during peak summer electricity consumption periods, the system assigns higher weights to paths passing through high-energy-consuming areas, guiding the selection of more energy-efficient routes. The principle behind this module is to abstract the physical path network into a mathematical graph theory problem, using a multi-objective optimization algorithm to find a comprehensive optimal solution under complex constraints.

[0137] The real-time linkage optimization module 25 implements a dynamic adjustment mechanism based on real-time feedback. This module continuously monitors equipment status and personnel location, and immediately initiates multi-device linkage analysis when a new fault signal is detected. For example, when the system detects two identical air conditioning units on the same floor reporting for repair simultaneously, it searches for nearby maintenance personnel with the corresponding skills. If it finds that Engineer Zhang is working in an adjacent area and is about to complete his task, the system marks the two fault points as "multi-device linkage" and recalculates the optimal path. During path optimization, the module calls a second multi-objective optimization function, adding a parts route coefficient. If a parts warehouse is found to be located within the path's sector area, a pickup point is forcibly inserted. The principle of this module lies in timely detection of optimization opportunities and dynamic adjustment of task allocation and path planning through real-time perception and predictive control.

[0138] The route distribution and verification module 26 completes the closed loop of solution execution and optimization. This module distributes optimized routes to the mobile terminals of maintenance personnel, while simultaneously updating the work arrangements of the parts warehouse and relevant departments. During execution, the module collects information such as actual route execution time, parts retrieval records, and customer feedback, and stores this information using blockchain technology. The verification module analyzes this data to evaluate the optimization effect of the integrated system of inventory, maintenance personnel, routes, and energy consumption. For example, by comparing the difference between actual energy consumption and predicted values, the system can adjust the energy consumption sensitivity calculation model; by analyzing task completion time deviations, the route weight calculation parameters can be optimized. The principle of this module lies in establishing a complete feedback learning mechanism, continuously optimizing the system decision model through continuous data collection and analysis, and achieving self-improvement and continuous improvement of the system.

[0139] This system, through the collaborative work of the aforementioned six modules, constructs a complete intelligent operation and maintenance decision-making system. In practical applications, such as the maintenance scenario of a central air conditioning system in a large industrial park, the system identifies multiple devices requiring maintenance through real-time monitoring. Through intelligent task dispatching and route optimization, tasks that originally required multiple individual operations are integrated into batch processing, improving operation and maintenance efficiency, reducing travel distance, and lowering energy consumption. This route planning system, based on digital twins and multi-objective optimization, effectively solves problems such as resource coordination difficulties and response delays in traditional operation and maintenance, achieving an intelligent transformation and upgrade of operation and maintenance management.

[0140] In one or more embodiments of this application, the verification module includes a blockchain evidence storage module. The blockchain evidence storage module adopts a one-time write + multiple verification mode. The on-chain data structure includes path hash, parts inventory change, energy consumption reduction estimate and customer receipt signature. The verification interface is opened to the energy diagnosis subsystem of the energy consumption management module. It completes the operation and maintenance-energy saving causal verification within a certain period of time without accessing the original work order. If the verification result shows that the energy consumption reduction is lower than the preset threshold, the system triggers a "path rollback" signal, so that the next round of planning will prioritize the use of historical high energy-saving path templates, thereby achieving a reliable closed-loop control of energy-saving effect.

[0141] The blockchain-based evidence storage function of the verification module adopts an innovative "write-once + verify-multiple times" model, with its core implementation based on distributed ledger technology and smart contract mechanisms. Upon completion of an operation and maintenance task, the system automatically packages key execution data into a data block containing path hashes, spare parts inventory change records, energy consumption reduction estimates, and customer electronic signatures, and writes it to the blockchain using a consensus algorithm. Specifically, the path hash is a unique identifier obtained by encrypting the spatiotemporal coordinate sequence of the actual execution path, ensuring the immutability of the path record; the spare parts inventory change record is linked to the real-time status update of the spare parts topology, forming a complete material flow trajectory; the energy consumption reduction estimate is calculated based on equipment operating data and energy consumption models, while the customer receipt signature uses digital certificate technology to ensure the authenticity of the feedback. These data, once written, form an immutable on-chain record, providing a reliable data foundation for subsequent verification and analysis.

[0142] The verification principle of this module is based on causal inference and time series analysis. After the verification interface is opened to the energy diagnostic subsystem, the system will automatically collect equipment energy consumption data within a preset time window (such as 24 hours after the completion of maintenance work). By comparing the changes in energy consumption under the same operating conditions before and after the operation, a causal relationship model between maintenance operations and energy-saving effects is established. For example, when a central air conditioning system completes heat exchanger cleaning and maintenance, the energy diagnostic subsystem will collect the unit's operating data under the same outdoor temperature and load rate for the next 24 hours and compare it with the data from the same period before maintenance. The entire process does not require access to the original work order details; verification can be completed solely through the path hash and energy consumption estimate stored on the blockchain, protecting business privacy while ensuring verification efficiency.

[0143] In one or more embodiments of this application, the path planning system further includes: an equipment management module, an equipment monitoring module, a microgrid module, an energy consumption management module, an operation and maintenance service module, and a customer management module; wherein, the multi-objective path planning module communicates in real time with the equipment management module, the equipment monitoring module, and the microgrid module, and the verification module communicates in real time with the energy consumption management module, the operation and maintenance service module, and the customer management module.

[0144] By establishing a real-time communication architecture between modules, cross-system data fusion and collaborative optimization were achieved. Real-time interaction between the multi-objective path planning module and the equipment management, equipment monitoring, and microgrid modules ensures that path planning comprehensively considers multiple factors such as equipment status, environmental parameters, and energy supply, thereby generating efficient and energy-saving operation and maintenance paths. Real-time data exchange between the verification module and the energy consumption management, operation and maintenance service, and customer management modules constructs a complete "execution-verification-optimization" closed loop, enabling the system to continuously improve path strategies based on real energy efficiency data and user feedback. This deeply integrated modular design breaks down the information silos of independent operation and maintenance in traditional systems, forming a data-driven, self-optimizing intelligent decision-making system, ultimately achieving the dual goals of improved operation and maintenance efficiency and reduced energy consumption.

[0145] See Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided according to an embodiment of this application. Figure 3 The electronic device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to execute the aforementioned multi-campus equipment maintenance work order dispatch method.

[0146] It should be understood that, in the embodiments of this application, the processor 301 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0147] Input device 302 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 303 may include a display (LCD, etc.), a speaker, etc.

[0148] The memory 304 may include read-only memory and random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include non-volatile random access memory. For example, the memory 304 may also store device type information.

[0149] In specific implementations, the processor 301, input device 302, and output device 303 described in the embodiments of this application can execute the implementation methods described in any embodiment of the equipment operation and maintenance work order dispatch method for multiple campuses provided in the embodiments of this application, or they can execute the implementation methods of the electronic devices described in the embodiments of this application, which will not be repeated here.

[0150] In another embodiment of this application, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program, which includes program instructions. When executed by a processor, the program instructions implement all or part of the processes in the multi-campus equipment maintenance work order dispatch method described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0151] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the foregoing embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device. Furthermore, the computer-readable storage medium can include both internal and external storage units of the electronic device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0152] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0153] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the electronic devices and units described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0154] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the couplings or direct couplings or communication connections shown or discussed may be indirect couplings or communication connections through some interfaces or units, or they may be electrical, mechanical, or other forms of connection.

[0155] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of this application, depending on actual needs.

[0156] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0157] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A multi-garden-oriented device operation and maintenance work order dispatching method, characterized in that, The method comprises the following steps: S100: constructing a multi-garden digital twin model, the digital twin model being configured as a "garden-building-space-equipment-fixture-maintenance personnel" six-level topology, and writing a dynamic attribute vector for each level of topology; S200: receiving a maintenance work order, extracting key features according to the maintenance work order, and splitting the maintenance work order into a feature vector including skill requirements, fixture requirements, and energy consumption sensitivity based on the key features; The extraction method of the key features in the step S200 includes: matching the corresponding skill requirement level from the equipment topology attribute based on the equipment type; matching the inventory status of the required fixture from the fixture topology attribute according to the fault type; and calculating the energy consumption influence coefficient of the maintenance operation in combination with the garden-building-space and equipment energy consumption level; S300: generating an initial dispatching scheme based on the six-level topology and the feature vector; The generation method of the initial dispatching scheme in the step S300 includes: taking "equipment-fixture-maintenance personnel" as the matching node in the six-level topology, monitoring the real-time inventory of the fixture warehouse, if the quantity of the required fixture is greater than the real-time inventory of the fixture warehouse, excluding the fixture corresponding to the equipment node, uploading the maintenance work order exception signal, and triggering a purchase reminder in reverse; if the quantity of the required fixture is less than or equal to the real-time inventory of the fixture warehouse, matching the maintenance personnel with corresponding skill authentication according to the skill requirements and the energy consumption sensitivity, forming the initial pairing of the equipment and the maintenance personnel, and obtaining the initial dispatching scheme; S400: performing preliminary path planning of the maintenance personnel according to the initial dispatching scheme; The generation method of the preliminary path planning of the maintenance personnel in the step S400 includes: taking the "garden-building-space" three-level topology points in the six-level topology as path vertices, taking the internal roads of the garden, the garage driveway, and the building elevator as edges, constructing a directed graph, and writing a real-time weight for each edge, the real-time weight being calculated by the vehicle flow density obtained from the equipment monitoring system, the real-time occupancy rate of the charging pile provided by the micro-grid system, and the energy consumption prediction curve obtained from the energy consumption management system; establishing a first multi-objective optimization function according to the directed graph, the first multi-objective optimization function including at least two of minimizing skill matching difference, minimizing fixture acquisition time, minimizing overall path planning time, and minimizing energy consumption influence coefficient; obtaining the preliminary path planning by solving the first multi-objective optimization function; S500: dynamically adjusting the initial dispatching scheme based on the real-time feedback parameters of the equipment monitoring system to obtain an optimized dispatching scheme while the maintenance personnel are executing the preliminary path planning. The parameter dynamic adjustment based on the real-time feedback of the equipment monitoring system in the step S500 obtains an optimized dispatching scheme, including: evaluating the equipment state and the operation and maintenance personnel state in real time according to the equipment monitoring system, the equipment state including at least one of new equipment failure and equipment failure priority, and the operation and maintenance personnel state including at least one of position change, work progress and load condition; when the required operation and maintenance of the new equipment failure is consistent with the current required operation and maintenance, marking the operation and maintenance as "multi-device linkage", and preferentially searching for other work order paths currently being executed, if an operation and maintenance personnel is about to enter the space where the new equipment failure is located and has the required skills to handle the new equipment failure, merging the next node of the operation and maintenance personnel into the current node to complete batch operation and maintenance, and generating an optimized dispatching scheme based on the equipment state and the operation and maintenance personnel state; S600: adjusting the preliminary path planning based on the optimized dispatching scheme to obtain an optimized path; S700: the operation and maintenance personnel performing operation and maintenance on the equipment based on the optimized path and the optimized dispatching scheme.

2. The method of claim 1, wherein, The step S600 of adjusting the preliminary path planning based on the optimized dispatching scheme to obtain an optimized path includes: after a new node marked as "multi-device linkage" is inserted, immediately taking the current operation and maintenance personnel position as the starting point, calling a "space-equipment-accessory" sub-topology in a six-level topology, recalculating the minimum cost path, and adding an accessory-in-order coefficient to the first multi-objective optimization function to obtain a second multi-objective optimization function, and if the next accessory warehouse is located in a preset sector of the line connecting the starting point and the new node, then forcibly taking the accessory in order; solving the second multi-objective optimization function to obtain an optimized path.

3. The method of claim 2, wherein, The step S600 of adjusting the preliminary path planning based on the optimized dispatching scheme to obtain an optimized path further includes: when the operation and maintenance personnel load rate exceeds a certain preset value and the path length exceeds the standard due to the insertion of a new node, enabling a load balancing sub-loop; the load balancing sub-loop is represented as searching for other operation and maintenance personnel in an idle state and having the same skill label in the six-level topology, if the straight-line distance between the current position of the operation and maintenance personnel and the new node is less than a certain proportion of the distance between the original operation and maintenance personnel and the new node, then migrating the new node and the subsequent in-order nodes to the operation and maintenance personnel in the idle state and having the same skill label, and regenerating a double path; taking the double path as the optimized path.

4. A path planning system for multi-park device operation and maintenance work orders, configured to perform the device operation and maintenance work order dispatching method of any one of claims 1-3. It includes: A digital twin module for constructing and maintaining a "park-building-space-equipment-accessory-operation and maintenance personnel" six-level topology, and writing a dynamic attribute vector in real time; A feature vector engine module for receiving operation and maintenance work orders, extracting skill requirements, accessory requirements and energy consumption sensitivity based on equipment type, fault type and park-building-space energy consumption level, and forming a calculable feature vector; An initial dispatching and inventory hard constraint module for excluding missing part nodes with real-time inventory of accessory storage as a hard constraint, and triggering a procurement reminder in reverse, and generating an initial dispatching scheme according to skill authentication and energy consumption sensitivity; The multi-objective path planning module is configured to construct a directed graph with the internal roads of the park, the garage lanes and the building elevators as edges, and to calculate the edge weights by combining the traffic flow density, the charging pile occupancy rate and the energy consumption prediction curve, and to output a preliminary path planning by solving a first multi-objective optimization function; The real-time linkage optimization module is configured to dynamically mark a "multi-device linkage" node and insert a current path based on the new fault signals of the device monitoring system, the personnel positions and the load states during the execution of the operation and maintenance personnel, and to obtain an optimized path by recalculating the minimum cost path by calling a second multi-objective optimization function; The path issuing and verification module is configured to issue the final optimized path to the operation and maintenance personnel, and to write the path results, the accessory taking records and the customer feedback information into a verification module, and the verification module is configured to verify whether the integration of the inventory, the operation and maintenance personnel, the path and the energy consumption is optimal.

5. The path planning system of claim 4, wherein, The verification module includes a blockchain storage module, which adopts a one-time writing + multiple verification mode, and the on-chain data structure includes a path hash, an accessory inventory change, an energy consumption reduction estimate and a customer receipt signature; the verification interface is open to an energy diagnosis subsystem of the energy consumption management module, and the operation and maintenance- energy saving causal verification is completed within a certain time without accessing the original work order; if the verification result shows that the energy consumption reduction is lower than a preset threshold, the system triggers a "path rollback" signal to make the next round of planning preferentially use a historical high energy saving path template, thereby realizing a credible closed-loop control of the energy saving effect.

6. The path planning system of claim 4, wherein, The path planning system further includes a device management module, a device monitoring module, a micro-grid module, an energy consumption management module, an operation and maintenance service module and a customer management module; the multi-objective path planning module communicates with the device management module, the device monitoring module and the micro-grid module in real time, and the verification module communicates with the energy consumption management module, the operation and maintenance service module and the customer management module in real time.

Images