New energy equipment maintenance scheme optimization system and method combined with working condition simulation

Abstract

The invention relates to the technical field of new energy equipment operation and maintenance management, in particular to a new energy equipment maintenance scheme optimization system and method combined with working condition simulation, and the system comprises a cloud optimization center, an edge execution unit and a field terminal which interact with each other through a secure communication link. The cloud optimization center fuses real-time working condition data (including operation, weather, resource data and historical maintenance records) of a new energy station, establishes a feature matrix, performs parallel simulation through a working condition simulation model and a digital twin model, identifies a maintenance opportunity window, generates a maintenance task package and optimizes a scheduling scheme. And the edge execution unit deploys a site, collects real-time data, executes emergency re-planning and safety interlocking control, and starts local simplified model emergency re-planning when communication is interrupted or working conditions are suddenly changed. And the on-site terminal receives the task package and the scheduling scheme, completes personnel identification, risk prompting and operation evidence obtaining, and returns operation data to the cloud end, so that self-learning and rolling updating of the digital twin model are realized.

Description

Technical Field

[0001] This invention relates to the field of new energy equipment operation and maintenance management technology, specifically to a new energy equipment maintenance scheme optimization system and method that combines operating condition simulation. Background Technology

[0002] With the continuous growth of installed capacity of new energy sources such as wind power, photovoltaics, and energy storage, the scale of power plants and the number of equipment are constantly expanding. The rationality and execution efficiency of equipment maintenance plans have a direct impact on the economic efficiency and safety of power generation. Traditional maintenance plans rely heavily on manual experience and are usually based on annual or quarterly cycles, lacking the ability to dynamically respond to weather, grid connection, resource availability, and equipment operating status.

[0003] Under complex weather conditions and uncertain operating conditions, the execution of maintenance plans often faces multiple challenges. On the one hand, disturbances from external factors such as wind speed, electricity prices, and grid-connected capacity can lead to inappropriate selection of maintenance windows, resulting in increased power generation losses or dispatch conflicts. On the other hand, insufficient on-site resource constraints, personnel skill matching, and accessibility of tools and equipment can easily cause delays in maintenance tasks and safety risks. In addition, traditional planning models cannot be quickly adjusted to emergencies, lacking the ability to tolerate grid outages and replan in emergencies, resulting in low reliability of plan execution.

[0004] While some existing technologies incorporate big data analytics and intelligent scheduling methods, most remain at the static optimization level, failing to combine operational simulation and digital twin models to achieve dynamic maintenance optimization. This makes it difficult to achieve closed-loop control of "prediction-decision-execution-feedback" in complex operating conditions and network environments. Therefore, there is an urgent need for a new energy equipment maintenance optimization system that combines operational simulation, dynamic optimization, and edge fault-tolerant control to improve the adaptability and safe execution capability of maintenance plans. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a system and method for optimizing the maintenance scheme of new energy equipment by combining working condition simulation, in order to address the shortcomings of the prior art. This system and method are used to solve the technical problems of static planning, delayed risk response and insufficient execution fault tolerance in the existing maintenance management of new energy equipment.

[0006] The objective of this invention is achieved through the following technical solutions: In a first aspect, the present invention provides a new energy equipment maintenance scheme optimization system that combines working condition simulation, including a cloud optimization center, an edge execution unit, and a field terminal; the cloud optimization center, the edge execution unit, and the field terminal interact through a secure communication link. The cloud-based optimization center is used to acquire real-time operating condition data from new energy power plants, perform fusion analysis, and establish a feature matrix. It then uses a pre-constructed operating condition simulation model and a digital twin model to perform parallel simulation of the feature matrix. Based on the simulation results, it identifies maintenance opportunity windows and generates maintenance task packages and optimized scheduling schemes, which are then distributed to field terminals. The real-time operating condition data includes acquired operational data, meteorological data, resource data, and historical maintenance records. The edge execution unit is deployed at the new energy equipment site to collect real-time operating condition data, perform emergency replanning and safety interlock control based on the real-time operating condition data, and initiate emergency replanning based on the local simplified model when communication is interrupted or operating conditions change suddenly. The field terminal is used to receive maintenance task packages and optimized scheduling plans, perform personnel identification, risk warnings and work verification based on the maintenance task packages and optimized scheduling plans, obtain work data, and transmit the work data back to the cloud optimization center to realize the self-learning and rolling update of the digital twin model.

[0007] As a further improvement of the present invention, the simulation and risk assessment module specifically includes: The multi-source data fusion module is used to uniformly collect and preprocess real-time operating data of new energy power stations, and to fuse and analyze the preprocessed real-time operating data to form a feature matrix. The real-time operating data includes the operating status of the power station equipment, meteorological environment, scheduling plan, personnel resources and historical maintenance records. The operating condition disturbance scenario generation module is used to construct several sets of disturbance scenario simulation sample sets in the prediction time domain based on the wind speed, electricity price, grid-connected capacity, resource latency and safety risk factors of new energy power plants, and to train the operating condition simulation model based on the disturbance scenario simulation sample sets. The digital twin model library includes equipment-level twin models, system-level twin models, and resource-level twin models. Equipment-level twin models are used to characterize the health status of equipment; system-level twin models are used to characterize the system's power generation capacity; and resource-level twin models are used to characterize resource accessibility. The simulation and risk assessment module is used to simulate and calculate the power generation loss, safety risk and resource delay indicators of different maintenance schemes under various disturbance scenarios based on the feature matrix, operating condition simulation model and digital twin model library, and form a joint criterion for loss and risk. The opportunity window identification module is used to identify maintenance opportunity windows based on the loss-risk joint criterion and preset safety constraints. The maintenance opportunity window is a maintenance time interval that meets the conditions that the power generation loss is below a threshold, the safety risk is controllable, and the grid connection capacity and meteorological conditions are compliant. The task package generation and resource mapping module is used to generate maintenance task packages based on equipment health index and maintenance opportunity window, and establish resource mapping relationship between tasks and personnel, vehicles and materials based on graph structure algorithm, and verify resource reachability; The multi-objective rolling optimization scheduling module is used to dynamically sort and allocate resources for the maintenance task package using a rolling time-domain planning algorithm, with the optimization objectives of power generation loss, safety risk and resource utilization, to generate an optimized scheduling scheme. The compliance verification and ticket output module is used to verify the compliance of the optimized scheduling scheme and generate electronic work orders and operation tickets with digital signatures.

[0008] As a further improvement of the present invention, the simulation and risk assessment module specifically includes: Based on the aforementioned feature matrix, operating condition simulation model, and digital twin model library, time-domain rolling simulations are performed in parallel under multiple disturbance scenarios to calculate risk data under each disturbance scenario; the risk data includes power generation loss, safety risk indicators, human resource expenditure, logistics costs, and delay risk losses. The risk data is weighted and summed according to the preset weighting coefficients to obtain the loss-risk joint comprehensive index under each disturbance scenario; Calculate the expected value and standard deviation of the loss-risk joint comprehensive index under all disturbance scenarios, and combine it with the robustness adjustment coefficient to form a loss-risk joint criterion.

[0009] As a further improvement of the present invention, the multi-objective rolling optimization scheduling module is specifically used for: The maintenance cycle is divided into multiple rolling periods. In each period, a multi-objective mixed integer linear programming model is established with the optimization objectives of minimizing power generation loss, minimizing safety risk, and maximizing resource utilization. A combination of linear weighting and hierarchical decision-making is used to solve the multi-objective mixed integer linear programming model, and a scheme fluctuation penalty term and a self-learning weight adjustment mechanism are introduced. When weather forecasts, resource status, or power grid dispatch information are updated, it triggers a re-optimization and dynamic adjustment of the plan for subsequent periods.

[0010] As a further improvement of the present invention, the edge execution unit includes: The on-site data aggregation module is used to collect and cache on-site equipment operation information, meteorological parameters, and personnel location information in real time; The simplified simulation evaluation module is used to quickly estimate the safety of the operation and the feasibility of the task based on a local simplified model and real-time monitoring data in the event of communication failure. The emergency replanning module is used to recalculate and adjust the task execution sequence based on the latest operating conditions when wind speed exceeds the limit or resources are blocked. The safety interlock control module is used to acquire cached data and monitor the field data aggregation module; The execution status feedback module is used to send the operation process data and equipment status back to the cloud optimization center.

[0011] As a further improvement of the present invention, the field terminal includes: The work order receiving and task display module is used to receive and display the electronic work order, operation steps and required tool list; The personnel identification and tool binding module is used to identify tools based on the equipment health index and defect level in the maintenance task package, and bind the successfully verified operators to the corresponding tools according to the optimized scheduling scheme; The step-by-step prompts and risk alerts module is used to provide real-time operation guidance based on the task flow and to issue alerts when high-risk environmental conditions are detected. The job evidence collection and process tracking module is used to automatically record images, videos and time information of key job steps, and form tamper-proof tracking data after hash encryption; The results submission and feedback reporting module is used to upload the actual execution data and environmental parameters to the cloud optimization center after the job is completed.

[0012] Secondly, the present invention provides a method for optimizing maintenance schemes for new energy equipment by combining operating condition simulation, used to implement the above-mentioned system for optimizing maintenance schemes for new energy equipment by combining operating condition simulation, comprising: The real-time operating data of new energy power plants are collected and processed by the cloud-based optimization center to form a feature matrix; Based on the feature matrix, the feature matrix is ​​simulated in parallel using the constructed working condition simulation model and digital twin model. Based on the working condition simulation results, maintenance opportunity windows are identified, and maintenance task packages and optimized scheduling schemes are generated and sent to the field terminals. Edge execution units deployed on-site collect real-time operating data, perform safety interlock control, and initiate emergency replanning based on a local simplified model in the event of communication interruption or sudden changes in operating conditions. The maintenance task package and optimized scheduling scheme are received and executed using the field terminal. Personnel identification, risk warning and operation certification are performed, and the operation data is transmitted back to the cloud optimization center to realize the self-learning of the digital twin model and the updating of the rolling time domain planning algorithm.

[0013] As a further improvement of the present invention, the method of using the constructed operating condition simulation model and digital twin model to perform parallel simulation of the feature matrix, identifying maintenance opportunity windows based on the operating condition simulation results, and generating maintenance task packages and optimized scheduling schemes includes: By calling the device-level twin model, system-level twin model, and resource-level twin model in the digital twin model, the health status of the equipment, the power generation capacity of the system, and the resource accessibility parameters can be obtained. Based on the aforementioned feature matrix, operating condition simulation model, and digital twin model library, the power generation loss, safety risk, and resource delay indicators of different maintenance schemes under various disturbance scenarios are simulated and calculated to form a joint loss-risk criterion. Used to select maintenance time windows that meet the conditions based on the loss-risk joint criterion and preset safety constraints; Based on the equipment health status and the maintenance time window, a maintenance task package is generated, and a resource mapping relationship between the task and personnel, vehicles, and materials is established based on a graph structure algorithm to verify resource reachability. This method uses a rolling time-domain programming algorithm to dynamically sort and allocate resources for the maintenance task package, with the optimization objectives of power generation loss, safety risk, and resource utilization, to generate an optimized scheduling scheme.

[0014] As a further improvement to the present invention, the specific process for forming the joint loss-risk criterion includes:

[0015] In the formula, Indicating a joint criterion for loss and risk, This represents the power generation loss during maintenance shutdowns, calculated using a digital twin model. Indicators representing safety risks to personnel and equipment. This indicates personnel scheduling and human resource expenses. This indicates the logistics costs of transporting goods and equipment. This indicates the risk of loss due to maintenance delays caused by weather fluctuations or external factors. , , , , These are the weighting coefficients for each indicator.

[0016] As a further improvement of the present invention, the optimal matching mapping of tasks and resources specifically includes: A bipartite graph is constructed using task nodes and resource nodes. The weights of the edges in the bipartite graph represent the cost of the task's adaptation to the resource in terms of skills, geographical distance, and arrival time. The cost matrix is ​​corrected by introducing task priority weights; The Hungarian algorithm is used to find the minimum cost matching of the corrected cost matrix, thus obtaining the optimal task-resource allocation relationship.

[0017] The beneficial effects of this invention are as follows: This invention provides a new energy equipment maintenance scheme optimization system that combines operating condition simulation. The cloud optimization center establishes a feature matrix by integrating and analyzing real-time operation data, meteorological data, resource data, and historical maintenance records of new energy power plants. It then uses an operating condition simulation model and a digital twin model to perform parallel simulation of this feature matrix, thereby accurately identifying maintenance opportunity windows and generating maintenance task packages and optimized scheduling schemes, achieving scientific control of maintenance timing and efficient allocation of resources. The edge execution unit is deployed at the equipment site, collecting real-time operating condition data while performing emergency replanning and safety interlock control. In the event of communication interruption or sudden changes in operating conditions, it initiates emergency replanning based on a local simplified model, effectively ensuring the safe operation of the equipment under emergency conditions. The field terminal receives maintenance task packages and optimized scheduling schemes, identifies personnel, provides risk warnings, and collects work certifications, generating work data that is sent back to the cloud optimization center. This drives the digital twin model to complete self-learning and rolling updates, forming a closed-loop optimization mechanism. The three interact through a secure communication link, working together to achieve end-to-end optimization from working condition simulation analysis to on-site operation execution. Compared with existing technologies, this significantly improves the matching degree between maintenance plans and actual working conditions, the timeliness of emergency response, and the overall adaptability of the system. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of a new energy equipment maintenance scheme optimization system that combines working condition simulation according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the functional modules of the cloud optimization center according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the functional modules of the edge execution unit according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the functional modules of the field terminal in an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives and technical solutions of this invention clearer and easier to understand, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. The specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention.

[0021] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings and specific embodiments. The described embodiments are only some embodiments of the present invention, and not all embodiments.

[0022] Example 1 This embodiment provides a new energy equipment maintenance scheme optimization system that combines operating condition simulation. Its features include a cloud optimization center, an edge execution unit, and field terminals; the cloud optimization center, edge execution unit, and field terminals interact via a secure communication link. The cloud-based optimization center is used to acquire real-time operating condition data from new energy power plants for fusion analysis and to establish a feature matrix. It then uses a pre-constructed operating condition simulation model and a digital twin model to perform parallel simulations of the feature matrix. Based on the simulation results, it identifies maintenance opportunity windows and generates maintenance task packages and optimized scheduling plans, which are then distributed to field terminals. Real-time operating condition data includes acquired operational data, meteorological data, resource data, and historical maintenance records. Edge execution units are deployed at the new energy equipment site to collect real-time operating condition data, perform emergency replanning and safety interlock control based on the real-time operating condition data, and initiate emergency replanning based on the local simplified model when communication is interrupted or operating conditions change suddenly. The field terminal is used to receive maintenance task packages and optimized scheduling plans. Based on the maintenance task packages and optimized scheduling plans, it performs personnel identification, risk warnings, and work verification to obtain work data. The work data is then transmitted back to the cloud optimization center to enable the self-learning and rolling updates of the digital twin model.

[0023] The working principle of this embodiment is as follows: Real-time operating condition data from multiple sources, including operational data, meteorological data, resource data, and historical maintenance records, is acquired through a cloud-based optimization center, fused and analyzed, and a feature matrix is ​​established. A pre-constructed operating condition simulation model and a digital twin model are used to perform parallel simulation of the feature matrix. Based on the simulation results, maintenance opportunity windows are identified, and maintenance task packages and optimized scheduling schemes are generated, thereby improving the comprehensiveness and timeliness of maintenance decisions. Edge execution units are deployed at the new energy equipment sites to collect real-time operating condition data. Based on this data, emergency replanning and safety interlock control are executed, and in the event of communication interruption or sudden changes in operating conditions, a local simplified model is used. Emergency replanning is initiated to enhance the real-time nature of on-site response and system robustness. After receiving maintenance task packages and optimized scheduling plans, on-site terminals identify personnel, provide risk warnings, and obtain work certifications to obtain work data. This work data is then transmitted back to the cloud optimization center to enable the self-learning and rolling updates of the digital twin model, thereby ensuring the safety traceability of the work process and the continuous optimization of the model. These technical features interact collaboratively through secure communication links to form a closed-loop system from data acquisition, simulation optimization, emergency control to feedback updates. This enables dynamic adaptive optimization of maintenance plans for new energy equipment, improving overall maintenance efficiency and safety.

[0024] Example 2 This invention provides a specific implementation of a new energy equipment maintenance scheme optimization system that combines operating condition simulation. This system aims to address problems existing in the maintenance of current new energy power plants, such as high power generation losses, unreasonable resource allocation, delayed operating condition response, and insufficient safety tolerance. By introducing operating condition simulation and digital twin models, it achieves intelligent generation, dynamic optimization, and safe execution control of maintenance schemes, thereby significantly improving the robustness and executability of maintenance plans.

[0025] like Figure 1 As shown, this system adopts a three-layer collaborative architecture of cloud-edge-device, including three functional layers: cloud optimization center, edge execution unit, and field terminal. The three layers achieve bidirectional transmission of data and instructions through a secure communication link, constructing a closed-loop system from data fusion to execution feedback.

[0026] Among them, the cloud optimization center is the core of the system's decision-making and analysis. It is responsible for integrating the operation data, meteorological data and resource information of new energy power stations, building digital twin models and performing operating condition simulation analysis. Through opportunity window identification and multi-objective rolling optimization algorithms, it generates maintenance plans that meet safety and economic constraints and outputs formal work orders and tickets.

[0027] Edge execution units are deployed at the site and are responsible for on-site situational awareness, simplified simulation evaluation, emergency replanning, and safety interlock control. When communication is interrupted or operating conditions change abruptly, the edge unit can independently execute fault-tolerant control logic to ensure that the maintenance process is safe and controllable.

[0028] The field terminal is an interactive device for operators, used to receive maintenance work orders, display operating steps, prompt risk information, record the operation process, and upload execution feedback, so as to realize the digitalization, traceability and closed-loop management of maintenance task execution.

[0029] The overall system operation flow is as follows: the cloud-based optimization center acquires multi-source heterogeneous data and establishes a feature matrix; based on the operating condition simulation results, it identifies the optimal maintenance window and generates a task package; the edge execution unit performs safety verification and emergency adjustments to the plan on-site; and the on-site terminal completes the work execution and certification according to the instructions. Finally, all on-site feedback information will be transmitted back to the cloud to update the digital twin model and continuously optimize subsequent maintenance plans.

[0030] Through such Figure 1 The three-level collaborative architecture shown in this invention realizes a closed-loop control logic of simulation prediction, optimization decision-making, execution feedback, and model update for maintenance plans. This system can dynamically schedule and robustly execute maintenance tasks under complex and variable new energy operating conditions, thereby minimizing power generation losses while ensuring safety and significantly improving the intelligence and efficiency of new energy equipment maintenance.

[0031] To facilitate understanding of the technical solution of the present invention, the following description is provided in conjunction with the appendix. Figure 2 To be continued Figure 4 This paper further explains the structural composition and working principle of each major functional level in the system of this invention. Specifically, based on the overall architecture, the system achieves closed-loop control of the entire process from data fusion and working condition simulation to task execution and feedback optimization through the functional division and collaborative operation of the cloud optimization center, edge execution units, and field terminals. Each functional unit is both independent and organically connected, jointly ensuring the accurate generation, real-time updating, and safe execution of maintenance plans.

[0032] like Figure 2 As shown, the cloud-based optimization center serves as the core decision-making layer of the system, enabling centralized management of data processing, simulation calculations, and scheme optimization within a cloud environment. The cloud-based optimization center sequentially includes a multi-source data fusion module, a working condition disturbance scenario generation module, a digital twin model library, a simulation and risk assessment module, an opportunity window identification module, a task package generation and resource mapping module, a multi-objective rolling optimization scheduling module, and a compliance verification and ticket output module.

[0033] The multi-source data fusion module is used to uniformly collect and preprocess heterogeneous data from multiple sources, including the operating status, meteorological environment, dispatching plans, personnel resources, and historical maintenance records of new energy power plants. The module's data input sources include real-time measurement signals from wind turbines, photovoltaic inverters, energy storage converters, and grid-side monitoring systems, as well as information on wind speed, wind direction, irradiance, humidity, temperature, and thunderstorm probability provided by meteorological monitoring devices. Simultaneously, the system also accesses management data from the dispatching platform, such as maintenance plans, electricity price curves, grid-connected power limits, personnel scheduling, and material inventory.

[0034] During data processing, the module first synchronizes the time of different sampling frequencies and communication delay signals using a timestamp alignment algorithm. Then, it uses a spatial index matching mechanism to complete the unified mapping of device coordinates, geographical location information, and field distribution model. Subsequently, the system performs outlier detection and noise filtering on the raw data, and uses moving median filtering and clustering anomaly removal algorithms to remove communication interference and sensor drift errors, ensuring data quality and consistency.

[0035] The processed data is organized into a unified feature matrix with multi-dimensional characteristics, including equipment health status parameters, environmental load characteristics, and resource constraint information. This feature matrix serves as the core input for the subsequent "operating condition disturbance scenario generation module" and "simulation and risk assessment module," used to construct an operating condition disturbance sample set and calculate the loss-risk response relationship of maintenance plans, thereby providing basic data support for the overall analysis in the cloud-based optimization center.

[0036] The operating condition disturbance scenario generation module is used to construct a multi-scenario simulation sample set for the maintenance process of new energy equipment under multi-dimensional external environment and operating status changes. The module uses the feature matrix output by the multi-source data fusion module as input and models and combines four major categories of disturbance factors: meteorology, grid connection, resources, and operation.

[0037] First, the module extracts meteorological disturbance factors, including wind speed change rate, wind direction shift angle, irradiance fluctuations, and temperature and humidity coupling effects. Based on the statistical characteristics of historical time series and prediction models, it generates meteorological evolution trajectories for multiple future periods. Second, regarding grid constraints, the module considers external influences such as grid-connected power limits, electricity price fluctuations, and dispatch plan adjustments, establishing a grid-side disturbance model to reflect changes in power feedback or power limiting conditions.

[0038] At the resource level, the module introduces constraint parameters such as personnel availability, traffic delay, material allocation cycle and construction window accessibility to simulate the spatiotemporal uncertainty of on-site resources; at the operation level, the module generates equipment availability disturbance samples based on the changing trend of equipment health index and failure probability distribution.

[0039] To achieve high coverage and low bias in sample generation, the system employs a strategy combining Latin hypercube sampling and Monte Carlo simulation to perform multidimensional random sampling and combination of the aforementioned disturbance factors. Each sample group represents a potential operating condition evolution path, encompassing the joint change sequence of meteorological, grid connection, and resource conditions during the maintenance cycle. During the generation phase, the system automatically eliminates combined samples that do not meet physical constraints, retaining only the set of disturbance scenarios that conform to safety and operational boundaries.

[0040] Finally, the operational disturbance scenario generation module outputs a disturbance sample matrix containing multiple time periods and variables, and transmits it to the simulation and risk assessment module. This module uses a parallel computing platform to achieve batch generation and hierarchical storage of samples to support parallel loading and dynamic retrieval during subsequent simulation calculations, significantly improving the efficiency and accuracy of simulation analysis.

[0041] The digital twin model library is used for virtual mapping and dynamic synchronization of new energy equipment, system operation, and resource allocation. It is a core module for achieving simulation calculation accuracy and adaptive rolling optimization capabilities. This model library is built in a layered structure within a cloud-based optimization center, comprising three levels: equipment-level twins, system-level twins, and resource-level twins.

[0042] The equipment-level twin establishes a mapping model based on the operating data and health parameters of key equipment in new energy power plants (such as wind turbines, photovoltaic inverters, energy storage converters, and step-up transformers). The model takes equipment monitoring signals as input, including features such as temperature, current, voltage, speed, vibration, and active power. It generates an equipment health index (HI) through multi-dimensional time series modeling and health assessment algorithms. The system uses dynamic Bayesian inference and a time-varying weighted model to predict the trend of the health index and determine the availability and failure probability of the equipment during different maintenance periods.

[0043] The system-level twin is used to reflect the overall power generation performance, power distribution, and dynamic energy flow characteristics of the power plant. This layer of the model aggregates the outputs of multiple device-level twins and combines them with grid-connected constraints and scheduling plans to form a system power prediction and power loss model. Through power flow analysis and operation optimization algorithms, the system-level twin can calculate power generation efficiency, inverter losses, and grid-connected load margin in real time, providing macroscopic-level operating boundary conditions for simulation and risk assessment modules.

[0044] Resource-level twins are used to model the dynamic accessibility of personnel, vehicles, materials, and operational resources related to maintenance activities. The model combines GIS geographic information and scheduling data to establish a resource-task mapping relationship, calculating the path accessibility, transportation delay, and availability windows of each resource unit at different time periods. When optimizing scheduling module calls, this layer of the model can automatically update the accessibility matrix based on task priority and resource location, thereby improving the accuracy and timeliness of maintenance plan execution.

[0045] During system operation, the digital twin model library is updated in real time through a cloud-edge collaboration mechanism. The cloud model is responsible for overall prediction and scenario simulation, while the edge execution unit periodically reports on-site status data, including equipment operation deviations, weather changes, and task progress information. After receiving feedback from the edge, the cloud model corrects the model weights in real time through a parameter self-calibration algorithm to ensure that the twin model's calculation results remain highly consistent with the actual on-site conditions.

[0046] Ultimately, this module outputs the calculation results of equipment-level, system-level, and resource-level twins to the simulation and risk assessment module, providing high-precision input parameters for joint loss-risk calculation. At the same time, it provides real-time availability constraints and operational boundary data for the multi-objective rolling optimization scheduling module, enabling the maintenance plan generation process to be refined, dynamic, and adaptive.

[0047] The simulation and risk assessment module is used to quantitatively evaluate the economy, safety, and reliability of maintenance plans for new energy equipment under multiple disturbance scenarios. It is a crucial link in the cloud-based optimization center that connects operational condition simulation and plan decision-making. The module takes the disturbance sample set output by the operational condition disturbance scenario generation module as input and combines it with multi-layered model data from the digital twin model library to perform parallel simulation calculations on the execution process of different maintenance plans.

[0048] In its implementation, the module first standardizes the input parameters for each disturbance scenario, including meteorological elements (wind speed, wind direction, irradiance, temperature, and humidity), grid connection conditions (power limits, electricity price curves, and scheduling plans), resource constraints (personnel accessibility and material path delays), and equipment availability parameters (health index and predicted failure rate). Based on these parameters, the module constructs a time-domain rolling simulation sequence and executes simulation calculations simultaneously under multiple disturbance scenarios using a parallel computing framework.

[0049] In each simulation scenario, the system calculates four key indicators: power generation loss, maintenance delay risk, personnel safety risk, and logistics costs. These indicators form a continuous distribution sequence over time, reflecting the dynamic changes during the maintenance task execution process. To comprehensively evaluate the performance of different schemes under multi-disturbance conditions, the system introduces a joint loss-risk evaluation index J, calculated as follows:

[0050] In the formula, This represents a combined loss-risk index, used to measure the overall robustness of a maintenance plan under multiple disturbance scenarios. This represents the power generation loss during maintenance shutdowns, calculated using a digital twin model. This represents a safety risk indicator for personnel and equipment, taking into account the hazard level of the on-site working environment and the duration of task exposure. This indicates personnel scheduling and human resource expenses. This indicates the logistics costs of transporting goods and equipment. This indicates the risk of loss due to maintenance delays caused by weather fluctuations or external factors. , , , , The weighting coefficients for each indicator are determined by the site operation and maintenance strategy and safety management standards.

[0051] Weighting coefficient The value range is determined by the system's self-learning module based on historical evaluation results, and is used to balance the trade-off between power generation economy and safety.

[0052] To further assess the robustness of the solution, the module calculates the statistical characteristics of the loss-risk index based on simulation results under all perturbation scenarios, forming a robustness criterion expression:

[0053] In the formula, Indicates the comprehensive index under all disturbance scenarios. The expected value is used to evaluate the average performance level of the solution; Indicators The standard deviation is used to reflect the variability of the solution's performance across different scenarios; This is a robustness adjustment coefficient used to balance the economic efficiency and stability of a project; when the robustness adjustment coefficient... When the value is too high, the system prioritizes stability; when the value is too low, the system prioritizes economy. This parameter is adaptively updated by the cloud model based on seasonal characteristics and maintenance risk level. This is a robustness threshold used to limit the acceptable range of scheme performance.

[0054] When the criterion is satisfied If the system determines that the current maintenance plan meets the robust requirements in terms of economy, safety, and resource coordination, then the system will automatically trigger a recalculation or mark the plan as suboptimal.

[0055] The module outputs its calculation results in matrix form, including the distribution of index values ​​under various disturbance scenarios, statistical summaries, and a candidate set of optimal solutions. The output is then passed to the opportunity window identification module to determine the maintenance time intervals that meet the constraints, providing input for subsequent multi-objective rolling optimization scheduling.

[0056] In terms of engineering implementation, the simulation and risk assessment module is deployed on a cloud server using a containerized microservice architecture, supporting parallel operation and dynamic task allocation. The system can automatically schedule computing resources according to the scale of the site and the real-time computing load, ensuring that large-scale scenario simulations are completed within a limited time. Through the above design, this module not only improves the real-time performance and accuracy of simulation calculations, but also achieves quantitative assessment and dynamic support for the robustness of maintenance decisions under complex external disturbances.

[0057] The opportunity window identification module is used to filter maintenance periods that meet robustness constraints and safety thresholds from the evaluation results output by the simulation and risk assessment modules under multi-perturbation scenarios. It is a key unit for realizing dynamic time-series optimization of maintenance plans. This module uses a joint loss-risk index... Using robustness criteria results as input, and combining meteorological forecasts, grid connection scheduling plans, and resource constraints, the system identifies suitable time windows for maintenance operations at new energy power plants.

[0058] During system operation, the module first receives multi-dimensional time-series data output by the simulation and risk assessment module, including information such as power generation loss curves, risk distribution sequences, and resource delay weights. The module then uses a sliding time window mechanism to perform segmented calculations on this data, obtaining comprehensive indicators for each candidate time slice. , With real-time safety margin.

[0059] To ensure the safety and economy of maintenance operations, the system performs the following judgments within each time slice:

[0060] The time interval is marked as an "optional maintenance window" only when the condition is met; if the threshold is exceeded, the system automatically removes the time period and proceeds to the next window for judgment.

[0061] Within the set of time intervals that meet the robustness requirements, the module further integrates meteorological, grid connection, and resource constraints for screening.

[0062] 1. Meteorological constraints: The module calculates the rate of change of wind speed, wind direction shift angle, and thunderstorm probability to ensure that the on-site environment meets the operational safety requirements; 2. Grid constraint level: The system verifies the grid connection plan, power limits and dispatch commands for this time period to ensure that maintenance operations will not cause system power flow to exceed limits or grid connection fluctuations; 3. Resource constraint level: By calling the reachability matrix output by the resource-level twin, the availability status of personnel, vehicles and materials within the candidate time window is determined.

[0063] The module employs a hierarchical multi-objective screening strategy, calculating the comprehensive fitness function for candidate windows that meet robustness criteria and constraints. Furthermore, a heuristic search method is used to sort and prioritize different windows to ensure the convergence and interpretability of the screening process.

[0064] Its main evaluation factors include power generation loss rate, safety risk weight, and resource arrival time deviation. The system assigns a weighted score to all feasible windows and sorts them from highest to lowest score, ultimately outputting the optimal maintenance window and a set of alternative windows.

[0065] In terms of engineering implementation, the opportunity window identification module operates based on an event-driven mechanism. When the simulation module completes a new round of rolling calculations or external input data is updated (such as weather forecast corrections or changes in dispatch orders), the module automatically triggers a re-screening process. The system can dynamically adjust the judgment cycle according to the window update frequency, achieving real-time response and adaptive adjustment to external disturbances.

[0066] Finally, the identified maintenance window information is transmitted to the task package generation and resource mapping module, serving as a time constraint for subsequent maintenance task orchestration and scheduling optimization. Through this process, the opportunity window identification module ensures that maintenance activities are carried out within time periods with sufficient safety margins, controllable power generation losses, and available resources, thereby achieving dynamic safety decision-making and optimal time allocation for the maintenance plan of new energy power plants.

[0067] The task package generation and resource mapping module is used to match the identified maintenance windows with the equipment defects, maintenance tasks and resource constraints to be handled in the site, generate executable maintenance task packages and establish task-resource mapping relationships. It is one of the core modules for realizing the structuring of maintenance plans and scheduling optimization.

[0068] During system operation, this module first receives the optimal and alternative maintenance time period information output by the opportunity window identification module, and combines it with the health index of the equipment-level twin in the digital twin model library. Based on failure probability data, all equipment is prioritized. The module categorizes tasks into three levels according to equipment health level, failure risk, and maintenance urgency: critical maintenance tasks, planned maintenance tasks, and conditional inspection tasks.

[0069] Subsequently, the module constructs an initial task set based on task level and time window constraints, and determines the energy impact range of the tasks according to the power distribution model provided by the system-level twin, avoiding the superposition of local power losses caused by multiple tasks running concurrently. For each candidate task, the system extracts the status information of corresponding personnel, vehicles, materials, and tools through the resource-level twin to form a task requirement vector, which includes task execution time, skill level requirements, types of tools required, transportation delay, and safety qualification conditions.

[0070] In the task-resource matching process, the module employs a graph-based resource mapping algorithm to construct a bidirectional task-resource association model. The system uses task nodes and resource nodes as the two sides of a graph, with edge weights representing task-resource suitability, comprehensively considering geographical distance, arrival time, and skill matching. The module uses a minimum-cost matching algorithm (Hungarian algorithm) to solve for the optimal matching relationship between tasks and resources. This algorithm introduces task priority weights and resource arrival time constraints on top of the traditional cost matrix, adjusting task importance through a dynamic weighting function to achieve rapid matching and improved result stability under multi-objective conditions, thereby enhancing the real-time performance and reliability of maintenance task allocation in multi-constraint scenarios.

[0071] To further ensure the feasibility of the solution, the module includes a resource accessibility verification mechanism. The system calculates the spatial path and traffic delay between the resource's starting point and the maintenance site based on a GIS geographic model, determining whether it falls within the planned time window. Whether the internal resources can arrive on time; if there is a resource delay or unreachable status, the task will be automatically marked as "needs to be rescheduled" and sent back to the scheduling module for re-optimization.

[0072] After resource mapping is complete, the module generates structured task packages, each containing the following information: Task identification and priority coding; Corresponding equipment number and location; Specify the time window and allowable delay range; List of required personnel, vehicles, materials and tools; Safe operating requirements and risk control measures.

[0073] After the generated task package is verified by the compliance verification module, it is directly passed to the multi-objective rolling optimization scheduling module for global task sorting and time optimization. Through the processing of this module, the originally scattered maintenance tasks are transformed into structured schedulable objects, realizing the digital description and precise resource allocation of maintenance operations at new energy power plants.

[0074] The multi-objective rolling optimization scheduling module serves as the decision-making core of the cloud-based optimization center. It is used to perform time sequencing, resource allocation, and scheme re-optimization of maintenance tasks under dynamic external disturbances. This module takes the structured task set output by the task package generation and resource mapping module as input, and combines real-time meteorological, resource availability, and power grid scheduling constraint information to establish a multi-objective rolling optimization model, thereby realizing the dynamic generation and continuous correction of maintenance schemes.

[0075] During the system initialization phase, the module defines an optimization target set based on the site operation strategy, including: 1. The objective of minimizing power generation losses is to reduce the impact of maintenance shutdowns on the overall system output; 2. The safety risk minimization objective is used to ensure the safety of personnel and equipment during maintenance; 3. The goal of maximizing resource utilization is used to improve the efficiency of human and material allocation; 4. The goal of task execution continuity is to reduce cross-task switching time and scheduling conflicts.

[0076] To simultaneously meet the above objectives, the system constructs a multi-objective optimization model and solves it using a combination of linear weighting and hierarchical decision-making. In the linear weighting stage, the module sets dynamic weight coefficients based on operating conditions and management priorities, constructing the optimization objective function in a comprehensive form. Subsequently, in the hierarchical decision-making stage, the system prioritizes meeting safety and power generation loss constraints, and then refines resource and timing optimizations while satisfying the first two conditions.

[0077] The module employs a hybrid solution approach combining mixed-integer linear programming (MILP) and heuristic search methods for rolling optimization computation. To improve the stability of the solution, a solution fluctuation penalty term is introduced into the MILP model to constrain the differences between solutions in adjacent rolling periods. Simultaneously, a self-learning weight adjustment mechanism is implemented, automatically correcting the target weights by analyzing historical execution feedback, thereby improving the continuity and adaptability of the optimization model. For large-scale task sets, the system further employs genetic algorithms or particle swarm optimization algorithms for global search to shorten the solution time and avoid local optima.

[0078] For large task sets, the system performs a global search using a genetic algorithm or particle swarm optimization (PSO) and solves for local optima using MILP, thereby obtaining a high-quality executable solution within a limited time.

[0079] This module operates within a rolling time domain structure, dividing the entire maintenance cycle into multiple rolling periods. Each period is optimized independently, and scheme splicing and continuity constraint verification are performed at the transition between periods. When weather forecasts, resource availability, or power grid dispatch information change, the system automatically triggers the rolling optimization process, adaptively adjusting the task sequence and resource allocation for subsequent periods, thereby achieving dynamic updates and real-time responses to the plan.

[0080] Furthermore, to avoid execution instability caused by frequent changes in the solution, the module sets a solution fluctuation penalty. The system compares the task ranking differences between adjacent optimization periods, and automatically increases the penalty weight when the fluctuation exceeds a set threshold to balance the optimality and stability of the solution. The module also has a "self-learning optimization weight mechanism," which statistically corrects the weights of different objectives based on execution feedback data collected during long-term operation, allowing the model to gradually approach the globally steady-state optimal strategy through multiple rounds of rolling.

[0081] Ultimately, the module outputs optimized results including structured information such as task execution sequence, resource scheduling table, personnel shift allocation table, and equipment maintenance time schedule. These results are then passed to the compliance verification and ticket output module for safety verification and work order generation. Through this design, the module achieves adaptive rolling optimization control of the maintenance plan, enabling the maintenance of new energy equipment to maintain efficient, safe, and stable operation even in complex environments and under multiple disturbances.

[0082] The compliance verification and ticketing output module is used to systematically verify the safety, operational consistency, and procedural compliance of maintenance plans after multi-objective rolling optimization, and to generate maintenance work orders and operation tickets with legal effect and on-site execution guidance. This module is located in the output layer of the cloud optimization center and is a key link in realizing the closed loop from plan decision-making to on-site execution.

[0083] In the specific implementation process, the module first receives the optimization scheme data output by the multi-objective rolling optimization scheduling module, including the task sequence table, personnel allocation table, equipment maintenance list, and resource scheduling plan. Based on the electrical primary wiring model, grid topology, and operation mode database, the system performs electrical topology consistency checks on the maintenance scheme, verifying whether the operation sequence conforms to the safety logic of circuit breaking, isolation, voltage testing, grounding, and power restoration. If a topology conflict or operational logic incompleteness is found, the system automatically alerts to the risk and returns the scheme to the next higher-level module for correction.

[0084] Subsequently, the module invokes the safety procedure knowledge base to verify each item in the plan, including the work content, work permit categories, and safety measures. The knowledge base includes national and industry standards, site-specific operating procedures, and equipment-level safety technical requirements. The system uses rule matching and semantic comparison to determine whether the plan complies with operating procedures and maintenance authorization requirements. Through knowledge base retrieval and automatic comparison mechanisms, the system effectively prevents situations such as unauthorized operation tickets, incorrect isolation points, or omissions of safety measures.

[0085] After verification, the module enters the ticket generation stage. The system automatically generates operation tickets, work tickets, and safety briefing sheets corresponding to the task package. Each ticket includes a task number, work time, personnel list, isolation measures, power restoration conditions, and emergency contact information. The generated tickets are stored in a standardized data format in a cloud database and pushed to the field terminal via a secure interface. After receiving the tickets at the field terminal, the operator must confirm their execution permission through an identity verification module (such as facial recognition or work permit scanning) before the system can authorize the operation.

[0086] To further ensure traceability, the module establishes an execution verification chain simultaneously with ticket generation. All work orders and tickets generate digital signature records at each stage of issuance, receipt, execution, completion, and archiving, forming a chain-like verification path. The cloud server encrypts and stores these signature records, ensuring the tamper-proof nature and audit traceability of ticket data.

[0087] During ticket execution, the system also supports real-time monitoring and status synchronization. When the edge execution unit or field terminal detects a change in task execution status (e.g., maintenance completed, grid connection restored, or safety measures lifted), the module automatically updates the ticket status and generates an execution closed-loop report in the cloud. The report includes execution time, records of operational risk events, and safety measure receipts, which are used for subsequent risk analysis and model self-learning updates.

[0088] Through the above design, the compliance verification and ticket output module achieves automatic conversion and full-process supervision from optimized solutions to execution work orders. The introduction of this module not only eliminates subjective errors in the manual review process, but also builds an electronic operation system that meets the safety production standards of new energy power plants, providing technical support for on-site safety control, process transparency, and traceability of responsibility.

[0089] like Figure 3 As shown, the edge execution unit is deployed at the new energy power station site and serves as the execution and fault tolerance layer of this system. It mainly includes a field data aggregation module, a simplified simulation evaluation module, an emergency replanning module, a safety interlock control module, and an execution status feedback module.

[0090] 1. On-site data aggregation module: This module collects on-site unit operation information, meteorological parameters, equipment sensor data and personnel location information in real time, and caches key indicators locally for rapid response to emergencies.

[0091] 2. Simplified Simulation Evaluation Module: In the event of cloud communication interruption or excessive network latency, the system quickly estimates the operational safety and task feasibility under the current conditions based on a local simplified model, providing a basis for subsequent emergency replanning. This simplified model is established based on a low-order linear approximation of the cloud twin model, using real-time monitoring data (wind speed, temperature, voltage) as input variables, and outputting a safety margin factor and a predicted task delay value, thus enabling rapid judgment even in a network outage environment.

[0092] 3. Emergency Replanning Module: This module recalculates the task execution sequence based on the latest on-site conditions. When wind speed exceeds the limit or resources are blocked, it automatically triggers delay or task switching strategies to ensure the continuity of the plan execution process.

[0093] 4. Safety Interlock Control Module: This module monitors wind speed, temperature and humidity, thunderstorm probability and hoisting stability in real time. When the monitoring results exceed the set threshold, the system automatically issues a stop command and records the timestamp, forming a closed loop for operational safety.

[0094] 5. Execution Status Feedback Module: This module feeds back operation process data, equipment status, and risk event records to the cloud in real time to enable the digital twin to learn and update its model.

[0095] like Figure 4 As shown, the field terminal is an intelligent interactive device used by operators, including a work order receiving and task display module, a personnel identification and tool binding module, a step prompt and risk alarm module, a work evidence collection and process recording module, and a result submission and feedback reporting module.

[0096] Work order receiving and task display module: The terminal receives work order and task information sent from the cloud via wireless communication and displays the maintenance objectives, operation steps and required tools.

[0097] Personnel identification and tool binding module: This module is used to verify the identity of operators through facial recognition or electronic certificates, and to bind and confirm tools to ensure consistency between safety responsibilities and operating permissions.

[0098] Step prompts and risk alarm module: The terminal displays operation prompts in real time according to the task flow, and issues alarm prompts when high wind speed, thunderstorms or mechanical abnormalities are detected.

[0099] Operation Evidence Collection and Process Traceability Module: During maintenance, the terminal automatically records images, videos, and time information of key steps to form complete operation traceability data, ensuring process traceability.

[0100] Before being uploaded, the trace data is encrypted with a local hash digest to generate a unique signature number, ensuring the immutability of the data during cloud comparison and auditing.

[0101] Result Submission and Feedback Reporting Module: After the task is completed, the terminal uploads data such as the actual execution time, environmental parameters, and security status to the cloud, providing support for the updating and rolling optimization of the digital twin.

[0102] This invention is achieved through, as follows Figures 1 to 4 The multi-layered collaborative architecture shown realizes the operational condition simulation prediction, dynamic optimization scheduling, edge fault-tolerant control, and execution feedback closed loop of the maintenance plan for new energy equipment. The cloud layer is responsible for intelligent decision-making, the edge layer is responsible for safe execution, and the terminal layer is responsible for interaction and feedback, forming a mutually supportive technical system. The system can achieve continuous, safe, and efficient execution of maintenance plans even under complex weather, resource fluctuations, and network instability environments, thereby effectively reducing power generation losses during the maintenance of new energy power plants and improving operational reliability and management intelligence.

[0103] In summary, the new energy equipment maintenance scheme optimization system proposed in this invention, which combines operating condition simulation, achieves a closed-loop intelligent control mechanism from operating condition prediction, scheme generation, on-site execution to data feedback by constructing a three-level collaborative architecture of cloud optimization center, edge execution unit, and field terminal. The system's overall design follows the technical route of "cloud-based global optimization, edge-based secure decision-making, and terminal-based precise execution," forming a comprehensive technical system of multi-source data fusion, simulation-driven decision-making, and fault-tolerant execution.

[0104] During system operation, the cloud-based optimization center dynamically maps the operating status of new energy equipment based on multi-source data fusion and digital twin modeling. Through disturbance scenario simulation and risk assessment, the system can accurately identify the optimal maintenance window within the prediction time domain. Combined with task package generation and multi-objective rolling optimization algorithms, it forms a maintenance scheduling scheme that balances safety and economy. Before being issued, the scheme undergoes dual compliance verification of power flow and ticketing to ensure the safe and executable nature of the output plan.

[0105] The edge execution unit, acting as the system's field control hub, can operate independently in the event of communication interruptions or unexpected situations. Its built-in simplified simulation model and emergency replanning logic can determine the risk level based on real-time monitoring data and automatically execute network outage fault-tolerant strategies, thereby ensuring the safety and continuity of maintenance operations in extreme environments. The safety interlock control module monitors key environmental parameters and personnel status in real time, immediately triggering shutdown and alarm mechanisms in the event of an anomaly, providing immediate safety assurance for on-site operations.

[0106] The field terminal serves as the system's human-machine interface, ensuring the digitalization and traceability of the entire maintenance process through functions such as visualized work orders, identity verification, risk alerts, and work certification. After the work is completed, the terminal transmits the actual working conditions and execution data back to the cloud, enabling dynamic correction and self-learning of the digital twin model, thus continuously improving the system's prediction accuracy and optimization capabilities.

[0107] Through the aforementioned three-level collaborative mechanism, this invention achieves a transformation in the maintenance of new energy equipment from "static planning" to "dynamic optimization driven by simulation," from "manual decision-making" to "model-based decision-making and adaptive execution," and from "one-way scheduling" to "closed-loop learning and feedback optimization." The system can automatically adjust the maintenance rhythm and resource allocation according to the operating conditions and meteorological characteristics of different sites, achieving a balance between safety, economy, and real-time performance in the maintenance plan.

[0108] The technical advantages of this invention are mainly reflected in the following aspects: 1. Improved robustness: Through disturbance scenario simulation and robustness assessment, the system can generate stable and reliable maintenance solutions under various uncertain conditions.

[0109] 2. Enhanced security: The edge execution unit implements multi-level safety interlocks and emergency replanning, effectively preventing risk events caused by sudden changes in operating conditions.

[0110] 3. Enhanced intelligence: The cloud-based optimization center uses digital twins and rolling optimization algorithms to enable self-learning and self-correction of maintenance plans, giving the system the ability to continuously evolve.

[0111] 4. Improved execution efficiency: The field terminal enables electronic task assignment and process tracking, reducing manual coordination and recording steps and significantly improving the execution efficiency of maintenance operations.

[0112] In summary, this invention deeply integrates operating condition simulation, digital twin, optimized scheduling and edge fault tolerance technology to construct an intelligent maintenance optimization system for new energy power plants. It can achieve safe, stable and efficient maintenance organization and execution in complex environments, and has significant engineering practical value and prospects for promotion and application.

[0113] Example 3 This embodiment provides a method for optimizing maintenance schemes for new energy equipment based on operating condition simulation. It is implemented using the new energy equipment maintenance scheme optimization system based on operating condition simulation described in Embodiments 1 and 2. The steps include: The real-time operating data of new energy power plants are collected and processed by the cloud-based optimization center to form a feature matrix; Based on the feature matrix, a working condition simulation model containing multiple sets of disturbance scenario sample sets is constructed. The feature matrix is ​​simulated in parallel using the constructed working condition simulation model and the digital twin model. Based on the working condition simulation results, maintenance opportunity windows are identified, and maintenance task packages and optimized scheduling schemes are generated. The maintenance task packages and optimized scheduling schemes are then sent to the field terminals. Edge execution units deployed on-site collect real-time operating data, perform safety interlock control, and initiate emergency replanning based on a local simplified model in the event of communication interruption or sudden changes in operating conditions. The maintenance task package and optimized scheduling scheme are received and executed using the field terminal. Personnel identification, risk warning and operation certification are performed, and the operation data is transmitted back to the cloud optimization center to realize the self-learning of the digital twin model and the updating of the rolling time domain planning algorithm.

[0114] Example 4 In another embodiment of the present invention, a computer-readable storage medium is provided as a storage component within a terminal device, the function of which is to store programs and data. It should be noted that the computer-readable storage medium here encompasses not only the built-in storage components of the terminal device but also extended storage components supported by the device. Essentially, it is a tangible medium capable of containing or storing programs that can be invoked by or in conjunction with an instruction execution system, device, or apparatus. This storage medium provides storage areas for the terminal's operating system and stores one or more instructions suitable for processor loading and execution, which can constitute one or more computer programs containing program code.

[0115] Specifically, examples of computer-readable storage media (a non-exclusive list) include: electrical connections with one or more wires, portable disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, optical fibers, portable optical disc read-only memory, optical storage devices, magnetic storage devices, or any reasonable combination of the above types.

[0116] The storage medium may also include data signals propagated as part of a baseband portion or a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any reasonable combination of both. Furthermore, computer-readable storage medium may also refer to other readable media besides conventional readable storage media, capable of sending, propagating, or transmitting programs for use or operation by an instruction execution system, apparatus, or device. Program code on the storage medium can be transmitted via any suitable medium, including but not limited to wireless, wired, optical fiber, or any reasonable combination thereof.

[0117] The program code used to implement the operations of this invention can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C. The execution modes of the program code include: running entirely on the user's computing device, running partially on the user's device as a standalone software package, running partially in a distributed manner on both the user's device and a remote computing device, or running entirely on a remote computing device or server. When a remote computing device is involved, the device can be connected to the user's computing device via any type of network such as a local area network (LAN) or a wide area network (WAN), or connected to an external computing device via the Internet through an Internet service provider.

[0118] The processor is capable of loading and executing one or more instructions stored in a computer-readable storage medium to implement the corresponding steps of the new energy equipment maintenance scheme optimization method combined with operating condition simulation described in Example 1.

Claims

1. A system for optimizing maintenance schemes for new energy equipment by combining operating condition simulation, characterized in that, It includes a cloud optimization center, an edge execution unit, and field terminals; the cloud optimization center, edge execution unit, and field terminals interact through a secure communication link; The cloud-based optimization center is used to acquire real-time operating condition data from new energy power plants, perform fusion analysis, and establish a feature matrix. It then uses a pre-constructed operating condition simulation model and a digital twin model to perform parallel simulation of the feature matrix. Based on the simulation results, it identifies maintenance opportunity windows and generates maintenance task packages and optimized scheduling schemes, which are then distributed to field terminals. The real-time operating condition data includes acquired operational data, meteorological data, resource data, and historical maintenance records. The edge execution unit is deployed at the new energy equipment site to collect real-time operating condition data, perform emergency replanning and safety interlock control based on the real-time operating condition data, and initiate emergency replanning based on the local simplified model when communication is interrupted or operating conditions change suddenly. The field terminal is used to receive maintenance task packages and optimized scheduling plans, perform personnel identification, risk warnings and work verification based on the maintenance task packages and optimized scheduling plans, obtain work data, and transmit the work data back to the cloud optimization center to realize the self-learning and rolling update of the digital twin model.

2. The new energy equipment maintenance scheme optimization system combining operating condition simulation as described in claim 1, characterized in that, The cloud-based optimization center includes: The multi-source data fusion module is used to uniformly collect and preprocess real-time operating data of new energy power stations, and to fuse and analyze the preprocessed real-time operating data to form a feature matrix. The real-time operating data includes the operating status of the power station equipment, meteorological environment, scheduling plan, personnel resources and historical maintenance records. The operating condition disturbance scenario generation module is used to construct several sets of disturbance scenario simulation sample sets in the prediction time domain based on the wind speed, electricity price, grid-connected capacity, resource latency and safety risk factors of new energy power plants, and to train the operating condition simulation model based on the disturbance scenario simulation sample sets. The digital twin model library includes equipment-level twin models, system-level twin models, and resource-level twin models. Equipment-level twin models are used to characterize the health status of equipment; system-level twin models are used to characterize the system's power generation capacity; and resource-level twin models are used to characterize resource accessibility. The simulation and risk assessment module is used to simulate and calculate the power generation loss, safety risk and resource delay indicators of different maintenance schemes under various disturbance scenarios based on the feature matrix, operating condition simulation model and digital twin model library, and form a joint criterion for loss and risk. The opportunity window identification module is used to identify maintenance opportunity windows based on the loss-risk joint criterion and preset safety constraints. The maintenance opportunity window is a maintenance time interval that meets the conditions that the power generation loss is below a threshold, the safety risk is controllable, and the grid connection capacity and meteorological conditions are compliant. The task package generation and resource mapping module is used to generate maintenance task packages based on equipment health index and maintenance opportunity window, and establish resource mapping relationship between tasks and personnel, vehicles and materials based on graph structure algorithm, and verify resource reachability; The multi-objective rolling optimization scheduling module is used to dynamically sort and allocate resources for the maintenance task package using a rolling time-domain planning algorithm, with the optimization objectives of power generation loss, safety risk and resource utilization, to generate an optimized scheduling scheme. The compliance verification and ticket output module is used to verify the compliance of the optimized scheduling scheme and generate electronic work orders and operation tickets with digital signatures.

3. The new energy equipment maintenance scheme optimization system combining operating condition simulation as described in claim 2, characterized in that, The simulation and risk assessment module specifically includes: Based on the aforementioned feature matrix, operating condition simulation model, and digital twin model library, time-domain rolling simulations are performed in parallel under multiple disturbance scenarios to calculate risk data under each disturbance scenario; the risk data includes power generation loss, safety risk indicators, human resource expenditure, logistics costs, and delay risk losses. The risk data is weighted and summed according to the preset weighting coefficients to obtain the loss-risk joint comprehensive index under each disturbance scenario; Calculate the expected value and standard deviation of the loss-risk joint comprehensive index under all disturbance scenarios, and combine it with the robustness adjustment coefficient to form a loss-risk joint criterion.

4. The new energy equipment maintenance scheme optimization system combining operating condition simulation as described in claim 2, characterized in that, The multi-objective rolling optimization scheduling module is specifically used for: The maintenance cycle is divided into multiple rolling periods. In each period, a multi-objective mixed integer linear programming model is established with the optimization objectives of minimizing power generation loss, minimizing safety risk, and maximizing resource utilization. A combination of linear weighting and hierarchical decision-making is used to solve the multi-objective mixed integer linear programming model, and a scheme fluctuation penalty term and a self-learning weight adjustment mechanism are introduced. When weather forecasts, resource status, or power grid dispatch information are updated, it triggers a re-optimization and dynamic adjustment of the plan for subsequent periods.

5. The new energy equipment maintenance scheme optimization system combining operating condition simulation as described in claim 1, characterized in that, The edge execution unit includes: The on-site data aggregation module is used to collect and cache on-site equipment operation information, meteorological parameters, and personnel location information in real time; The simplified simulation evaluation module is used to quickly estimate the safety of the operation and the feasibility of the task based on a local simplified model and real-time monitoring data in the event of communication failure. The emergency replanning module is used to recalculate and adjust the task execution sequence based on the latest operating conditions when wind speed exceeds the limit or resources are blocked. The safety interlock control module is used to acquire cached data from the field data aggregation module and monitor environmental safety parameters in real time. When the monitored value exceeds the set threshold, it automatically triggers a safety pause command. The execution status feedback module is used to send the operation process data and equipment status back to the cloud optimization center.

6. The new energy equipment maintenance scheme optimization system combining operating condition simulation as described in claim 2, characterized in that, Field terminals include: The work order receiving and task display module is used to receive and display the electronic work order, operation steps and required tool list; The personnel identification and tool binding module is used to identify tools based on the equipment health index and defect level in the maintenance task package, and bind the successfully verified operators to the corresponding tools according to the optimized scheduling scheme; The step-by-step prompts and risk alerts module is used to provide real-time operation guidance based on the task flow and to issue alerts when high-risk environmental conditions are detected. The job evidence collection and process tracking module is used to automatically record images, videos and time information of key job steps, and form tamper-proof tracking data after hash encryption; The results submission and feedback reporting module is used to upload the actual execution data and environmental parameters to the cloud optimization center after the job is completed.

7. A method for optimizing maintenance schemes for new energy equipment by combining operating condition simulation, used to implement the new energy equipment maintenance scheme optimization system by combining operating condition simulation as described in any one of claims 1 to 6, characterized in that, include: The real-time operating data of new energy power plants are collected and processed by the cloud-based optimization center to form a feature matrix; Based on the feature matrix, the feature matrix is ​​simulated in parallel using the constructed working condition simulation model and digital twin model. Based on the working condition simulation results, maintenance opportunity windows are identified, and maintenance task packages and optimized scheduling schemes are generated and sent to the field terminals. Edge execution units deployed on-site collect real-time operating data, perform safety interlock control, and initiate emergency replanning based on a local simplified model in the event of communication interruption or sudden changes in operating conditions. The maintenance task package and optimized scheduling scheme are received and executed using the field terminal. Personnel identification, risk warning and operation certification are performed, and the operation data is transmitted back to the cloud optimization center to realize the self-learning of the digital twin model and the updating of the rolling time domain planning algorithm.

8. The method for optimizing maintenance schemes for new energy equipment based on operating condition simulation according to claim 7, characterized in that, The process involves using a pre-constructed operating condition simulation model and a digital twin model to perform parallel simulation of the feature matrix, identifying maintenance opportunity windows based on the simulation results, and generating maintenance task packages and optimized scheduling schemes, including: By calling the device-level twin model, system-level twin model, and resource-level twin model in the digital twin model, the health status of the equipment, the power generation capacity of the system, and the resource accessibility parameters can be obtained. Based on the aforementioned feature matrix, operating condition simulation model, and digital twin model library, the power generation loss, safety risk, and resource delay indicators of different maintenance schemes under various disturbance scenarios are simulated and calculated to form a joint loss-risk criterion. Used to select maintenance time windows that meet the conditions based on the loss-risk joint criterion and preset safety constraints; Based on the equipment health status and the maintenance time window, a maintenance task package is generated, and a resource mapping relationship between the task and personnel, vehicles, and materials is established based on a graph structure algorithm to verify resource reachability. This method uses a rolling time-domain programming algorithm to dynamically sort and allocate resources for the maintenance task package, with the optimization objectives of power generation loss, safety risk, and resource utilization, to generate an optimized scheduling scheme.

9. The method for optimizing the maintenance scheme of new energy equipment by combining working condition simulation according to claim 8, characterized in that, The specific process of forming a joint loss-risk criterion includes: In the formula, Indicating a joint criterion for loss and risk, This represents the power generation loss during maintenance shutdowns, calculated using a digital twin model. Indicators representing safety risks to personnel and equipment. This indicates personnel scheduling and human resource expenses. This indicates the logistics costs of transporting goods and equipment. This indicates the risk of loss due to maintenance delays caused by weather fluctuations or external factors. , , , , These are the weighting coefficients for each indicator.

10. The method for optimizing the maintenance scheme of new energy equipment by combining working condition simulation according to claim 8, characterized in that, The optimal matching mapping between tasks and resources specifically includes: A bipartite graph is constructed using task nodes and resource nodes. The weights of the edges in the bipartite graph represent the cost of the task's adaptation to the resource in terms of skills, geographical distance, and arrival time. The cost matrix is ​​corrected by introducing task priority weights; The Hungarian algorithm is used to find the minimum cost matching of the corrected cost matrix, thus obtaining the optimal task-resource allocation relationship.

Images