New energy equipment group-oriented operation and maintenance optimization strategy generation method, device and medium

Abstract

The application provides a new energy equipment group-oriented operation and maintenance optimization strategy generation method, equipment and medium, multi-source data of new energy equipment is collected to construct a unified feature table; risk and performance evaluation is carried out on new energy candidate equipment to be maintained based on the unified feature table, and the candidate equipment is sorted according to the maintenance priority score, the sorted candidate equipment is optimized and solved under multi-objective and multi-constraint conditions, a preliminary strategy containing preventive maintenance and corrective maintenance is generated, cost-benefit analysis is carried out on the preliminary strategy, a scheduling strategy is generated based on the analyzed preliminary strategy to perform resource scheduling, and hierarchical and time-sharing scheduling and rolling rearrangement are carried out under resource constraints to form an executable task set and a timing schedule; the application quantifies multi-dimensional indexes by maintenance priority calculation, combines multi-objective trade-off and active operation and maintenance closed-loop mechanism, and carries out dynamic scheduling optimization under multiple resource constraints, thereby effectively reducing operation and maintenance cost.

Description

Technical Field

[0001] This invention relates to the field of operation and maintenance management and intelligent optimization technology for new energy equipment, specifically a method, equipment and medium for generating operation and maintenance optimization strategies for new energy equipment groups. Background Technology

[0002] With the rapid development of the new energy industry, the installed capacity of wind power, photovoltaic, and energy storage equipment continues to expand, and their operation and maintenance management faces multiple challenges, including massive data volume, wide equipment distribution, and complex operating environments. Existing operation and maintenance systems typically collect equipment operation data through SCADA systems, combined with manual inspection records and alarm events for status monitoring. In terms of maintenance decision-making, they mainly adopt a model that combines regular inspections with reactive maintenance, formulating maintenance plans based on equipment manufacturer recommendations or the experience of operation and maintenance personnel. In terms of resource scheduling, manual dispatching is based on shift schedules and spare parts inventory information.

[0003] However, the above methods have the following shortcomings in scenarios with multiple devices, multiple constraints, and strong uncertainty: insufficient utilization of multi-source data, lack of a unified risk measurement and maintenance priority ranking mechanism; weak ability to balance multiple objectives, making it difficult to simultaneously optimize costs, downtime, reliability improvement, and security risks; disconnect between prediction results and maintenance execution, making it difficult to transform predictive information into proactive operation and maintenance actions; lack of dynamic optimization in resource scheduling, resulting in low efficiency of cross-team collaboration; and the fact that most systems are open-loop architectures, making it impossible to achieve adaptive optimization of strategies through execution feedback. Summary of the Invention

[0004] This invention provides a method for generating operation and maintenance optimization strategies for new energy equipment clusters. By uniformly quantifying multi-dimensional indicators through maintenance priority calculation, combining multi-objective trade-offs and proactive operation and maintenance closed-loop mechanisms, and performing dynamic scheduling optimization under multiple resource constraints, the method improves the accuracy of operation and maintenance decisions and resource utilization efficiency, effectively reduces operation and maintenance costs, and solves the problems in the background technology.

[0005] To achieve the above objectives, the technical solution of the present invention is as follows: A method for generating operation and maintenance optimization strategies for new energy equipment clusters involves executing the following steps using computer equipment: S1. Collect multi-source data from new energy equipment and construct a unified feature table; the multi-source data includes SCADA data, maintenance records, alarm events, environmental and traffic information, grid-connected production constraints, and cost data; S2, based on a unified feature table, performs risk and performance assessment on candidate new energy equipment to be maintained, obtains equipment-level risk score, urgency, impact, cost and constraint penalty, then calculates maintenance priority, and sorts candidate equipment according to maintenance priority score; S3, under multiple objectives and constraints, optimize the ranked candidate devices to generate a preliminary strategy that includes preventive maintenance and corrective maintenance. S4. Perform cost-benefit analysis on the preliminary strategy, generate a scheduling strategy based on the preliminary strategy after analysis, and perform hierarchical time-sharing scheduling and rolling rescheduling under the resource constraints of personnel, spare parts, tools and vehicles to form an executable task set and time sequence schedule. Preferably, the formula for calculating the maintenance priority score in S2 is:

[0006] Where i represents the device. This represents the maintenance priority score for device i. To score risk, For urgency, For the degree of influence, For cost, To restrain punishment, Risk scoring The corresponding weights Risk scoring The corresponding weights Risk scoring The corresponding weights Risk scoring The corresponding weights Risk scoring The corresponding weights; Risk scores, urgency, impact, and costs are all normalized using a distributed adaptive method before calculation; constraints and penalties are based on factors such as grid-connected production window matching degree, geographical distance and route accessibility, personnel skill matching degree, and spare parts availability. When the same priority score occurs, the ranking is adjusted according to the preset decision mutual exclusion rules and cooperation rules. The cooperation rules include the joint maintenance rule and the adjacent equipment linkage rule. The maintenance priority score also supports parallel alternative implementation methods, including: a nonlinear combination method that applies a monotonically nonlinear transformation to each indicator and then performs a weighted summation; a piecewise linear rule method that sets different slopes and intercepts for threshold, saturation and penalty segments; and a hierarchical rule priority method that first performs hard constraint screening based on safety and compliance rules and then sorts the remaining candidates.

[0007] Preferably, the weights in the priority score calculation formula in S2 are jointly calibrated by Pareto front selection and historical strategy effects, and are updated on a rolling basis with execution feedback; The calibration process includes: based on historical strategy effects and simulation results, adopting a hierarchical target weighting and adaptive update method, incorporating weights into a configuration repository for version management, and supporting one-click rollback.

[0008] Preferably, in S3, the specific steps for optimizing the ranked candidate devices under multiple objectives and constraints are as follows: S31, Determine the objectives and full constraints of the multi-objective optimization, where the optimization objectives include minimizing maintenance costs, minimizing downtime, maximizing reliability improvement, and minimizing security risks; Constraints include personnel or team skills and schedules, spare parts inventory and delivery time, geographical or transportation time, grid connection or production window, contract or warranty terms, as well as constraints on parallelism and mutual exclusion, task decomposition hierarchy, and upper limit constraints on waiting penalties and rearrangement costs. S32, select the optimization solution algorithm, adopt the core idea of ​​hierarchical objective trade-off and Pareto front selection, and combine integer linear programming, mixed integer linear programming, tabu search, genetic algorithm, simulated annealing, local search, large neighborhood search, variable neighborhood search or any combination thereof to carry out the solution; S33, Set solution constraints, limit the upper limit of computation time and the maximum number of iterations, and clarify the criteria for judging the near-optimal quality boundary. The quality boundary includes the ratio of the relative optimal value not exceeding a preset threshold, the absolute difference not exceeding a preset threshold, and the gap rate relative to the relaxation lower bound not exceeding a preset threshold. S34, perform phased solution, first construct a feasible solution within the set time boundary, and continuously improve and optimize the quality of the feasible solution; S35, trigger early termination judgment: when the improvement of the quality of the feasible solution is lower than the preset threshold, or when the solution process reaches the upper limit of the calculation time limit, the solution process is terminated immediately. S36, output the solution results and supporting evaluation information, output the optimized solution that satisfies the near-optimal quality boundary, and simultaneously output the quality evaluation and proof indicators, including the boundary estimate, gap rate, validation set evaluation score and solution trajectory summary.

[0009] Preferably, after generating the initial strategy including preventive and corrective maintenance in S3, the process also includes generating specific procedures for proactive maintenance actions: Based on the initial strategy to identify high-risk devices, preventive actions are generated from the preset strategy library. The preventive actions are selected by ranking by cost-benefit ratio and matching with scene templates. The proactive maintenance actions are linked to maintenance priorities to ensure the efficient coordination of preventive and corrective maintenance. The proactive maintenance actions include parameter calibration, scheduling of downtime or load reduction windows, lubrication or cleaning, early replacement of key components and pre-positioning of spare parts. Preferably, the cost-benefit analysis of the preliminary strategy in S4 specifically includes: cumulative cost, benefit, investment payback period and marginal benefit indicators; Preferably, the hierarchical time-sharing scheduling and rolling reordering in S4 specifically include: when the spare parts delivery time changes, the grid connection window is shortened, or the triggering condition of traffic obstruction occurs, the scheduling strategy is rolled and reordered, and the overall reordering cost is kept within a preset threshold. The scheduling strategy includes prioritizing nearby tasks, merging batch tasks, and optimizing routes, while also considering the overlap of time periods and skill matching for cross-shift collaboration. Resource scheduling is carried out in a hierarchical and time-sharing manner under the resource constraints of personnel, spare parts, tools and vehicles, so as to realize cross-team collaborative execution; Preferably, the unified feature table constructed in S1 includes: unique equipment identifier, key component name, observation decision timestamp, predicted failure probability, performance degradation degree, impact range measurement, key performance indicator loss ratio, expected maintenance downtime, current spare parts inventory, spare parts delivery date, team skill vector, team available working hours, grid connection load reduction allowable window, expected driving time and warranty contract terms status fields, covering equipment identifier, component information, time sequence information, risk performance indicators, resource constraints and core operation and maintenance parameters.

[0010] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0011] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0012] As can be seen from the above technical solution compared with the prior art, the present invention has the following beneficial effects: 1. This invention quantifies multiple indicators such as risk, impact, and cost through maintenance priority calculation, and combines multi-objective optimization and Pareto front selection to achieve a systematic trade-off between cost, downtime, reliability, and safety risks, outputting the optimal maintenance strategy. At the same time, it introduces an active operation and maintenance mechanism to transform predictive results into preventive actions, and achieves continuous self-optimization of the model and weights through closed-loop feedback, significantly improving the rationality, interpretability, and adaptability of strategy generation.

[0013] 2. This invention enhances resource scheduling efficiency and engineering implementation value. Under multiple resource constraints such as personnel, spare parts, tools, and vehicles, it achieves cross-team collaborative execution and dynamic optimization through hierarchical time-sharing scheduling and rolling rescheduling. This effectively shortens work order turnaround time, reduces operation and maintenance costs, and vehicle mileage. At the same time, it is compatible with domestic information technology stacks, supports localized deployment and multi-scenario expansion, and ensures system security, compliance, and reproducibility while increasing power generation and reducing unit energy costs. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the method steps of the present invention; Figure 2 This is a schematic diagram of the implementation process of an embodiment of the present invention. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0016] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate the present invention, but should not be used to limit the scope of the present invention.

[0017] This invention provides a method for generating operation and maintenance optimization strategies for new energy equipment clusters, such as... Figure 1 As shown, perform the following steps using a computer device: S1. Collect multi-source data from new energy equipment and construct a unified feature table; the multi-source data includes SCADA data, maintenance records, alarm events, environmental and traffic information, grid-connected production constraints, and cost data; S2, based on a unified feature table, performs risk and performance assessment on candidate new energy equipment to be maintained, obtains equipment-level risk score, urgency, impact, cost and constraint penalty, then calculates maintenance priority, and sorts candidate equipment according to maintenance priority score; S3, under multiple objectives and constraints, optimize the ranked candidate devices to generate a preliminary strategy that includes preventive maintenance and corrective maintenance. S4. Perform cost-benefit analysis on the preliminary strategy, generate a scheduling strategy based on the preliminary strategy after analysis, and perform hierarchical time-sharing scheduling and rolling rescheduling under the resource constraints of personnel, spare parts, tools and vehicles to form an executable task set and time sequence schedule. Example: like Figure 2As shown in this embodiment, a large onshore wind farm is used as an example. The wind farm has an installed capacity of 150 MW and a total of 25 6.0 MW wind turbine generators, which have been in operation for 5 years. The wind farm is equipped with a SCADA system to collect the unit operation data in real time. The operation and maintenance team consists of 12 people, divided into 3 shifts: mechanical, electrical, and comprehensive. It is equipped with 2 operation and maintenance vehicles, and the spare parts warehouse is located in the center of the farm area.

[0018] The main operational challenges currently faced include: a wide distribution of equipment, covering an area of ​​approximately 30 square kilometers, with complex road conditions including mountainous sections; strict power grid dispatch requirements, allowing only two periods of downtime per day: 09:30-11:30 and 13:00-15:00; procurement cycles for some imported spare parts reaching up to 30 days; and significant seasonal fluctuations in wind conditions, with downtime losses during the spring windy season being significantly higher than in summer. Based on these scenarios, this application employs a method for generating operational optimization strategies for new energy equipment clusters to conduct intelligent operational optimization.

[0019] The implementation process begins with data collection and feature construction. The system has initiated a new round of operation and maintenance strategy generation. The collected data includes: SCADA data: Active power, spindle temperature, gearbox vibration, generator speed and other measurement data sampled every 10 minutes over the past 30 days; Maintenance records: 12 maintenance work orders in the past 3 months, including fault codes, handling measures, working hours, replacement parts, etc.; alarm events; environmental and traffic information: daily weather forecast, road traffic status, vehicle travel time matrix for each machine position; grid connection constraints; cost data.

[0020] After data cleaning, feature engineering and multi-source alignment, a unified feature table at the device level was constructed, which includes 25 devices and 12 core feature fields for each device.

[0021] Risk assessment and prioritization: Based on a unified feature table, risk and performance assessments are conducted for each device.

[0022] Taking device WT-023 as an example: Risk Score R: Based on the integrated model, the output of spindle temperature trend, vibration amplitude, and alarm frequency is 0.78; Urgency U: Due to the occurrence of 5 "spindle over-temperature" alarms in the past 3 days, with a positive trend slope, the urgency score is 0.70; Impact I: This equipment is the central unit of the site, and if it is shut down, it will affect the reactive power support of the regional power grid, so the impact score is 0.65; Cost C: The estimated cost of replacing the spindle bearing, for example, labor cost of 8,000 yuan + spare parts cost of 42,000 yuan + downtime loss calculated as 3 hours × 6,000 kW × 0.4 yuan / kWh = 7,200 yuan, totaling 57,200 yuan, is normalized to 0.41 after being normalized by the historical cost distribution of the site; Constraint Penalty Pen: The penalty value is 0.28, which is calculated based on the grid connection window matching degree, geographical distance, personnel skills, and spare parts accessibility.

[0023] The maintenance priority score is calculated using the following formula:

[0024] Where i represents the device. This represents the maintenance priority score for device i. To score risk, For urgency, For the degree of influence, For cost, To restrain punishment, Risk scoring The corresponding weights Risk scoring The corresponding weights Risk scoring The corresponding weights Risk scoring The corresponding weights Risk scoring The corresponding weights; Similarly, the other 24 devices were calculated and sorted from highest to lowest score. WT-023, WT-017, and WT-009 ranked in the top three and entered the preliminary candidate set. For example, WT-017 scored 0.512 and WT-009 scored 0.496.

[0025] Multi-objective optimization solution; S31: Determine the optimization objective and total constraints Optimization objectives: minimize maintenance costs, minimize downtime, maximize reliability improvement, minimize safety risks; constraints: team skills, team schedule, spare parts inventory, delivery time, geographical location of each machine position, grid connection window, parallelism constraints; WT-023 and WT-009 can be merged for inspection and task decomposition levels. S32: Select the optimization algorithm, adopt hierarchical objective trade-offs and Pareto front selection, combine mixed integer linear programming, and call the CBC solver.

[0026] S33: Set solution constraints, maximum computation time: 300 seconds, maximum number of iterations: 1000; quality boundary: relative gap ratio ≤ 2%; S34-S35: Phased solution and early termination. Solution process: The first feasible solution (single equipment sequential maintenance plan) is obtained at 45 seconds. An improved solution (combined maintenance plan) is obtained at 128 seconds. After 215 seconds, the quality improvement rate is less than 0.001 for 60 consecutive rounds, triggering early termination.

[0027] S36: Output the solution results, output the Pareto front solution set, and select the comprehensive optimal solution. Task 1: Replacement of WT-023 spindle bearing (proactive maintenance: parameter calibration + lubrication); Task 2: WT-009 blade inspection (to be performed in conjunction with WT-023); Task 3: Offline oil sample testing of WT-017 gearbox (due to the lack of spare parts, it has been converted to monitoring). It also outputs a quality assessment report: the target value of the optimal solution, the lower bound, the gap ratio, and a summary of the solution trajectory.

[0028] Proactive maintenance actions are generated for WT-023 based on the initial strategy: Identify high-risk components – spindle bearings (predicted failure probability 0.62, degradation degree 0.58); Match preventative actions from the strategy library—optional actions include: (A) parameter calibration (spindle temperature control threshold optimization), (B) lubrication (adding grease), and (C) early replacement; Cost-benefit ratio ranking: Action A: cost 0.1 million yuan, benefit 0.8 million yuan (extended lifespan), ratio 8.0; Action B: cost 0.3 million yuan, benefit 1.2 million yuan, ratio 4.0; Action C: cost 5.7 million yuan, benefit 3.5 million yuan, ratio 0.61. Based on the scenario template (spindle bearing mid-risk + nearing replacement cycle), select the "Action A + Action B" combination.

[0029] Prioritizing coordinated maintenance, merging proactive maintenance actions with corrective maintenance mainframe replacements to ensure collaborative execution, cost-benefit analysis and explanation, and analysis of the WT-023 comprehensive strategy: Explainable output information: Risk reduction ΔR: -0.18, Downtime reduction ΔT: -2.5 hours, Revenue increase ΔB: +42,000 yuan, Cost ΔC: +4,000 yuan; Priority score contribution breakdown: R contributes 0.27, U contributes 0.14, I contributes 0.16, C contributes -0.05, and Pen contributes 0.02, forming a visual contribution pie chart.

[0030] Resource scheduling and rolling reordering; Collect resource information – Mechanical Group (3 people, skill MEC3), Electrical Group (2 people, skill ELE2), Comprehensive Group (2 people, skill INSP1); Daily shift schedule (Mechanical Group has tasks in the morning, free in the afternoon); Spare parts inventory (2 spindle bearings); Vehicles (2 vehicles, in normal condition); Traffic time matrix (18 minutes from WT-023 to WT-009).

[0031] Prioritizing proximity and merging batches – WT-023 and WT-009 are geographically close, and after merging, the total travel time is optimized from 42+38 minutes to 42+18 minutes (one round trip), saving 20 minutes.

[0032] Generate an executable plan; Morning: One person from the mechanical team and one person from the electrical team will perform the WT-009 blade inspection (electrical coordination required); Afternoon: The mechanical team will perform WT-023 spindle replacement (including proactive maintenance). Vehicle 1 was fully supported throughout the process, and one spare main shaft bearing has been dispatched from the warehouse; S4A5: Rolling reordering triggered. Warehouse notification indicates that the special lubricating grease, originally scheduled for delivery this afternoon, is delayed until the next day. The system initiates rolling reordering: the Pen value for lubrication actions in WT-023 proactive maintenance is recalculated and increased. The scheduler postponed the lubrication operation to be performed separately the next day, merging it with another nearby device, WT-011; The spindle replacement proceeded normally in the afternoon, and the rearrangement cost was calculated to be 2.8%, which is less than the 5% threshold. Closed-loop update; based on historical strategy effects, 12 maintenance operations in the past 3 months, and the results of this simulation, Pareto frontier selection is used to jointly calibrate the weights, for example: The value was adjusted from 0.35 to 0.34 due to improved accuracy in risk model predictions. The value was adjusted from 0.20 to 0.19, improving the timeliness of emergency alarm response. The impact assessment was adjusted from 0.25 to 0.26, improving the consistency between the assessment and the actual losses. The value was adjusted from 0.12 to 0.13. Adjusted from 0.08 to 0.08; The new weights have been added to the configuration repository, version number v2.1, while version v2.0 is retained for rollback support.

[0033] After triggering the rolling update, the system will use the new weights for the next round of policy generation.

[0034] Results verification: After the execution of this operation and maintenance strategy: Robustness tests showed that the Top-3 strategy achieved a stability of 90% under a weight ±10% perturbation scenario, validating the effectiveness and engineering applicability of the method.

[0035] This application quantifies multiple indicators such as risk, impact, and cost through maintenance priority calculation, and combines multi-objective optimization and Pareto front selection to achieve a systematic trade-off between cost, downtime, reliability, and security risks, outputting the optimal maintenance strategy. At the same time, it introduces an active operation and maintenance mechanism to transform predictive results into preventive actions, and achieves continuous self-optimization of the model and weights through closed-loop feedback, significantly improving the rationality, interpretability, and adaptability of strategy generation.

[0036] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0037] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0038] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the above embodiments of the operation and maintenance optimization strategy generation method for new energy equipment groups.

[0039] It is understood that the systems, devices, and storage media provided in the embodiments of the present invention correspond to the methods provided in the embodiments of the present invention, and the explanations, examples, and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

[0040] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.

[0041] For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access, or a data storage device such as a server or data center that integrates one or more available media.

[0042] The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid state disks (SSDs)).

[0043] It should be noted that in this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0044] Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0045] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0046] The embodiments of the present invention are given for the purposes of illustration and description. Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for generating operation and maintenance optimization strategies for new energy equipment clusters, characterized in that, Perform the following steps using a computer device: S1. Collect multi-source data from new energy equipment and construct a unified feature table; the multi-source data includes SCADA data, maintenance records, alarm events, environmental and traffic information, grid-connected production constraints, and cost data; S2, based on a unified feature table, performs risk and performance assessment on candidate new energy equipment to be maintained, obtains equipment-level risk score, urgency, impact, cost and constraint penalty, then calculates maintenance priority, and sorts candidate equipment according to maintenance priority score; S3, under multiple objectives and constraints, optimize the ranked candidate devices to generate a preliminary strategy that includes preventive maintenance and corrective maintenance. S4 performs a cost-benefit analysis on the preliminary strategy, generates a scheduling strategy based on the analyzed preliminary strategy to schedule resources, and performs hierarchical time-sharing scheduling and rolling rescheduling under the resource constraints of personnel, spare parts, tools and vehicles to form an executable task set and time sequence schedule.

2. The method for generating operation and maintenance optimization strategies for new energy equipment clusters as described in claim 1, characterized in that: The formula for calculating the priority score in S2 is as follows: Where i represents the device. This represents the maintenance priority score for device i. To score risk, For urgency, For the degree of influence, For cost, To restrain punishment, Risk scoring The corresponding weights Risk scoring The corresponding weights Risk scoring The corresponding weights Risk scoring The corresponding weights Risk scoring The corresponding weights; The risk score, urgency, impact, and cost are all normalized using a distribution adaptive method before calculation. The constraints and penalties take into account factors such as the matching degree of grid-connected production window, geographical distance and route accessibility, personnel skill matching, and spare parts availability. When the same priority score occurs, the ranking is adjusted according to the preset decision mutual exclusion rules and cooperation rules. The cooperation rules include the joint maintenance rule and the adjacent equipment linkage rule. The maintenance priority score also supports parallel alternative implementation methods, including: a nonlinear combination method that applies a monotonically nonlinear transformation to each indicator and then performs a weighted summation; a piecewise linear rule method that sets different slopes and intercepts for threshold, saturation, and penalty segments; and a hierarchical rule priority method that first performs hard constraint screening based on safety compliance rules and then sorts the remaining candidates.

3. The method for generating operation and maintenance optimization strategies for new energy equipment clusters as described in claim 2, characterized in that: The weights in the priority score calculation formula in S2 are jointly calibrated by Pareto front selection and historical strategy effects, and are updated on a rolling basis with execution feedback. The calibration process includes: based on historical strategy effects and simulation results, adopting a hierarchical target weighting and adaptive update method, incorporating weights into a configuration repository for version management, and supporting one-click rollback.

4. The method for generating operation and maintenance optimization strategies for new energy equipment clusters as described in claim 1, characterized in that: In S3, the specific steps for optimizing the ranked candidate devices under multiple objectives and constraints are as follows: S31, Determine the objectives and full constraints of the multi-objective optimization, where the optimization objectives include minimizing maintenance costs, minimizing downtime, maximizing reliability improvement, and minimizing security risks; Constraints include personnel or team skills and schedules, spare parts inventory and delivery time, geographical or transportation time, grid connection or production window, contract or warranty terms, as well as constraints on parallelism and mutual exclusion, task decomposition hierarchy, and upper limit constraints on waiting penalties and rearrangement costs. S32, select the optimization solution algorithm, adopt the core idea of ​​hierarchical objective trade-off and Pareto front selection, and combine integer linear programming, mixed integer linear programming, tabu search, genetic algorithm, simulated annealing, local search, large neighborhood search, variable neighborhood search or any combination thereof to carry out the solution; S33, Set solution constraints, limit the upper limit of computation time and the maximum number of iterations, and clarify the criteria for judging the near-optimal quality boundary. The quality boundary includes the ratio of the relative optimal value not exceeding a preset threshold, the absolute difference not exceeding a preset threshold, and the gap rate relative to the relaxation lower bound not exceeding a preset threshold. S34, perform phased solution, first construct a feasible solution within the set time boundary, and continuously improve and optimize the quality of the feasible solution; S35, trigger early termination judgment: when the improvement of the quality of the feasible solution is lower than the preset threshold, or when the solution process reaches the upper limit of the calculation time limit, the solution process is terminated immediately. S36, output the solution results and supporting evaluation information, output the optimized solution that satisfies the near-optimal quality boundary, and simultaneously output the quality evaluation and proof indicators, including the boundary estimate, gap rate, validation set evaluation score and solution trajectory summary.

5. The method for generating operation and maintenance optimization strategies for new energy equipment clusters as described in claim 4, characterized in that: After generating the initial strategy that includes preventive and corrective maintenance in S3, the process also includes generating specific proactive maintenance actions: Based on the initial strategy to identify high-risk devices, preventive actions are generated from the preset strategy library. The preventive actions are selected by ranking by cost-benefit ratio and matching with scene templates. The proactive maintenance actions are linked to maintenance priorities to ensure the efficient coordination of preventive and corrective maintenance. The proactive maintenance actions include parameter calibration, scheduling of downtime or load reduction windows, lubrication or cleaning, early replacement of key components and pre-positioning of spare parts.

6. The method for generating operation and maintenance optimization strategies for new energy equipment clusters as described in claim 1, characterized in that: The cost-benefit analysis of the preliminary strategy in S4 specifically includes: cumulative cost, benefit, investment payback period, and marginal benefit indicators.

7. The method for generating operation and maintenance optimization strategies for new energy equipment clusters as described in claim 6, characterized in that: The hierarchical time-sharing scheduling and rolling reordering in S4 specifically include: when the spare parts delivery time changes, the grid connection window is shortened, or the triggering condition of traffic obstruction occurs, the scheduling strategy is rolled and reordered, and the overall reordering cost is kept within a preset threshold. The scheduling strategy includes near-end priority, batch task merging and route optimization, and takes into account the time overlap and skill matching degree of cross-shift collaboration. The resource scheduling is carried out in a hierarchical and time-sharing manner under the resource constraints of personnel, spare parts, tools and vehicles, so as to realize cross-team collaborative execution.

8. The method for generating operation and maintenance optimization strategies for new energy equipment clusters as described in claim 1, characterized in that: The unified feature table constructed in S1 includes: unique equipment identifier, key component name, observation decision timestamp, predicted failure probability, performance degradation degree, impact range measurement, key performance indicator loss ratio, expected maintenance downtime, current spare parts inventory, spare parts delivery date, team skill vector, team available working hours, grid connection load reduction allowable window, expected driving time and warranty contract terms status fields, covering equipment identifier, component information, time sequence information, risk performance indicators, resource constraints and core operation and maintenance parameters.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it causes the processor to perform the steps of the method as described in any one of claims 1 to 8.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the computer program is executed by the processor, it causes the processor to perform the steps of the method as described in any one of claims 1 to 8.

Images