New energy power station intelligent operation and maintenance scheduling and resource optimization method and system

Abstract

The invention discloses an intelligent operation and maintenance scheduling and resource optimization method and system for a new energy power station, and belongs to the technical field of data processing and management.The method comprises the steps that real-time power generation data, equipment sensor data, weather forecast data, a power grid scheduling instruction and electricity market electricity price information of the new energy power station are obtained; a unified data matrix is generated through space-time alignment processing to execute bidirectional feedback prediction, and a corrected power generation plan is generated; a cooperative scheduling decision is generated through a multi-target dynamic balance algorithm, and a personnel dispatching scheme and a material allocation path are generated through real-time path planning; and generating a historical decision data set based on the execution result and the actual execution deviation, and adjusting prediction parameters and scheduling parameters through an incremental learning model. A closed-loop optimization method combining data fusion, collaborative prediction, multi-target scheduling and adaptive learning is adopted, global collaborative optimization of multiple factors such as power generation, operation and maintenance, energy storage and markets can be achieved, and the economic benefit and the intelligent level of overall operation of a power station are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing and management, and in particular to a method and system for intelligent operation and maintenance scheduling and resource optimization of a new energy power station. Background Art

[0002] As a vital component of modern power systems, new energy power plants' core business is to generate electricity using renewable energy sources such as wind and solar power, and to ensure safe, stable, and economical operation through intelligent operations and maintenance. With the expansion of power plant scale and the deepening of power market reforms, effectively managing the massive amounts of data within and outside the power plant and, based on this, achieving coordinated optimization across multiple business processes such as power generation planning, equipment maintenance, energy storage scheduling, and market transactions has become key to enhancing the commercial competitiveness of power plants.

[0003] In existing technologies, the operation and management of new energy power plants typically rely on multiple independent software systems. For example, the data acquisition and monitoring control system collects real-time operating data, the power forecasting system provides power generation forecasts based on weather forecasts, and the work order management system schedules operations and maintenance tasks. When making scheduling decisions, operators often first develop a power generation plan and then schedule maintenance based on the plan and equipment status. The decision-making process for each link is relatively independent, lacking unified optimization goals and information exchange. The allocation of operations and maintenance resources also relies heavily on manual assignment, which limits scheduling efficiency and response speed.

[0004] However, data standards and timescales vary between systems, creating data silos and making it difficult to form a unified data view that supports global decision-making. Secondly, power forecasting is mostly open-loop, failing to fully integrate the real-time health status of equipment. This leads to deviations between the forecast results and actual power generation capacity, and an inability to self-correct based on historical deviations. More importantly, the decision-making links of power generation, operation and maintenance, and energy storage are separated from each other, often leading to resource conflicts and offsetting benefits. For example, unnecessary shutdowns and maintenance at high electricity prices, or increased power generation losses due to untimely operation and maintenance responses, all of which limit the improvement of the overall economic benefits of power plants. Summary of the Invention

[0005] To solve the above problems, the present invention provides an intelligent operation and maintenance scheduling and resource optimization system for new energy power stations. It adopts a closed-loop optimization method that combines data fusion, collaborative prediction, multi-objective scheduling and adaptive learning. It can achieve global collaborative optimization of multiple factors such as power generation, operation and maintenance, energy storage and market, and improve the economic benefits and intelligence level of the overall operation of the power station.

[0006] The above objectives can be achieved through the following solutions: A method for intelligent operation and maintenance scheduling and resource optimization of a new energy power station comprises obtaining real-time power generation data, equipment sensor data, weather forecast data, grid dispatch instructions, and electricity market price information of the new energy power station, and generating a unified data matrix through spatiotemporal alignment processing; performing bidirectional feedback prediction based on the unified data matrix, the bidirectional feedback prediction comprising generating long-term and short-term prediction results, and using the short-term prediction results to perform error correction on the long-term prediction results to generate a revised power generation plan; obtaining equipment maintenance urgency and energy storage charge state, and combining the revised power generation plan with the electricity market price information to generate a coordinated scheduling decision comprising a power station output plan, energy storage charging and discharging strategies, and operation and maintenance task allocation instructions through a multi-objective dynamic trade-off algorithm; obtaining technician location information and spare parts inventory location information based on the operation and maintenance task allocation instructions, and generating a personnel dispatch plan and material allocation path through real-time path planning; recording the execution results of the coordinated scheduling decision and calculating the actual execution deviation, generating a historical decision data set based on the execution results and the actual execution deviation, and using the historical decision data set to adjust prediction parameters and scheduling parameters through an incremental learning model.

[0007] Optionally, generating a unified data matrix includes: acquiring real-time power generation data of new energy power stations, equipment sensor data, weather forecast data, power grid dispatch instructions and power market electricity price information to obtain a basic data set; extracting timestamps and geographic coordinate information from the basic data set; synchronizing the data in the basic data set to a unified time step based on the timestamp, and mapping the data in the basic data set to a unified spatial reference based on the geographic coordinate information to obtain a processed basic data set; and structurally integrating the processed basic data set to generate a unified data matrix.

[0008] Optionally, the method further includes: extracting temperature parameters and vibration parameters from the equipment sensor data; analyzing the temperature parameters and vibration parameters using preset health status thresholds to generate equipment health status; and calculating a failure risk coefficient as the equipment maintenance urgency based on the equipment health status and the time rate of change of the temperature parameters and vibration parameters.

[0009] Optionally, generating a revised power generation plan includes: extracting weather trends and historical power generation patterns from the unified data matrix, and performing long-term forecasts to generate long-term forecast results; extracting real-time equipment status data from the unified data matrix, and performing short-term forecasts to generate short-term forecast results; calculating the prediction error between the long-term forecast results and the short-term forecast results, and using the prediction error to perform back-propagation calibration on the long-term forecast results to generate a revised power generation plan.

[0010] Optionally, generating a short-term prediction result includes: extracting real-time equipment status data from the unified data matrix; correcting the real-time equipment status data using the equipment maintenance urgency; performing a short-term prediction based on the corrected real-time equipment status data to generate a short-term prediction result.

[0011] Optionally, the generation of a collaborative scheduling decision including a power plant output plan, energy storage charge and discharge strategies, and operation and maintenance task allocation instructions includes: obtaining the equipment maintenance urgency and energy storage charge status; constructing a multi-objective function with power generation efficiency, equipment life, grid demand, and economy as optimization objectives; using the revised power generation plan, the equipment maintenance urgency, the energy storage charge status, and the power market electricity price information as input variables of the multi-objective function; solving the multi-objective function to obtain an optimal balance solution, and resolving the optimal balance solution into a power plant output plan, energy storage charge and discharge strategies, and operation and maintenance task allocation instructions to obtain a collaborative scheduling decision.

[0012] Optionally, the generation of personnel dispatch plans and material allocation paths includes: parsing the operation and maintenance task allocation instructions to determine the required technician skills and spare parts types; obtaining technician location information and spare parts inventory location information; combining the technician location information and the technician skills to screen out candidate technicians, and combining the spare parts inventory location information and the spare parts type to screen out candidate spare parts libraries; combining the candidate technicians and the candidate spare parts libraries, and performing path optimization calculations, selecting a combination based on the total response time, and generating personnel dispatch plans and material allocation paths.

[0013] Optionally, the adjustment of prediction parameters and scheduling parameters includes: recording the execution results of the collaborative scheduling decision and calculating the actual execution deviation, generating a historical decision data set based on the execution results and the actual execution deviation; inputting the historical decision data set into the incremental learning model for training to identify systematic deviation patterns; generating deviation adjustment amounts based on the systematic deviation patterns; and applying the deviation adjustment amounts to update prediction parameters and scheduling parameters.

[0014] Optionally, the method further includes: judging whether a sudden drop in power generation will occur in the future based on the revised power generation plan; if so, triggering an emergency response, calculating a predicted power gap based on the revised power generation plan, and updating the energy storage charging and discharging strategy according to the power gap; synchronously adjusting the operation and maintenance task allocation instructions, and generating adaptive scheduling instructions in combination with the updated energy storage charging and discharging strategy.

[0015] Based on the same inventive concept, the present invention also provides an intelligent operation and maintenance scheduling and resource optimization system for a new energy power station, the system comprising: a data fusion module for acquiring real-time power generation data, equipment sensor data, weather forecast data, grid scheduling instructions and power market price information of the new energy power station, and generating a unified data matrix through spatiotemporal alignment processing; a collaborative prediction module for performing bidirectional feedback prediction based on the unified data matrix, the bidirectional feedback prediction including generating long-term prediction results and short-term prediction results, and using the short-term prediction results to perform error correction on the long-term prediction results to generate a revised power generation plan; an intelligent scheduling module for acquiring equipment maintenance urgency and energy storage charge status, and combines the revised power generation plan with the electricity market price information to generate a collaborative scheduling decision including the power station output plan, energy storage charging and discharging strategy, and operation and maintenance task allocation instructions through a multi-objective dynamic trade-off algorithm; a resource optimization module is used to obtain the technician location information and spare parts inventory location information based on the operation and maintenance task allocation instructions, and generate a personnel dispatch plan and material allocation path through real-time path planning; an adaptive learning module is used to record the execution results of the collaborative scheduling decision and calculate the actual execution deviation, generate a historical decision data set based on the execution results and the actual execution deviation, and use the historical decision data set to adjust the prediction parameters and scheduling parameters through an incremental learning model.

[0016] Compared with the prior art, the present invention has the following advantages: 1. This invention establishes a data fusion and closed-loop learning mechanism to align and deeply integrate scattered power plant data, meteorological information, and market signals in time and space. It also uses an incremental learning model to continuously correct systematic deviations, significantly improving prediction accuracy and the reliability of decision-making, providing a high-quality data foundation for refined operation and management of power plants. 2. This invention proposes a global collaborative optimization scheduling method. Through a multi-objective dynamic trade-off algorithm, it integrates multiple mutually constrained management objectives, such as power generation efficiency, equipment lifespan, grid demand, and economic efficiency, into a unified decision-making framework. This method achieves deep coupling and integrated scheduling of power generation, operation and maintenance, energy storage, and other services, overcoming the local optimality and global suboptimality problems caused by traditional step-by-step decision-making, and comprehensively improving the overall operational efficiency of power station assets. 3. The present invention realizes the dynamic and precise allocation of operation and maintenance resources. By analyzing the operation and maintenance requirements in the collaborative scheduling instructions and combining the real-time location information and path planning of personnel and materials, it generates the optimal personnel dispatch and material allocation plan, which greatly shortens the fault response and repair time, reduces the operation and maintenance costs and power generation losses, and enhances the agility and economy of power station operation management.

[0017] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 It is a flow chart of a method for intelligent operation and maintenance scheduling and resource optimization of a new energy power station according to an embodiment of the present invention.

[0020] Figure 2 Schematic diagram of the effect of continuously optimizing prediction accuracy according to an embodiment of the present invention.

[0021] Figure 3 It is a structural diagram of an intelligent operation, maintenance, scheduling and resource optimization system for a new energy power station according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0023] Reference Figure 1 One embodiment of the present invention proposes a method for intelligent operation and maintenance scheduling and resource optimization of a new energy power station. It adopts a closed-loop optimization method that combines data fusion, collaborative prediction, multi-objective scheduling and adaptive learning. It can achieve global collaborative optimization of multiple factors such as power generation, operation and maintenance, energy storage and market, and improve the economic benefits and intelligence level of the overall operation of the power station.

[0024] The method of this embodiment specifically includes: Acquire real-time power generation data from new energy power stations, equipment sensor data, weather forecast data, grid dispatch instructions, and power market price information, and generate a unified data matrix through spatiotemporal alignment processing; Based on the unified data matrix, performing bidirectional feedback forecasting, the bidirectional feedback forecasting including generating a long-term forecast result and a short-term forecast result, and using the short-term forecast result to perform error correction on the long-term forecast result to generate a revised power generation plan; Obtaining the equipment maintenance urgency and energy storage charge state, and combining the revised power generation plan with the electricity market price information, generating a collaborative scheduling decision including the power station output plan, energy storage charging and discharging strategy, and operation and maintenance task allocation instructions through a multi-objective dynamic trade-off algorithm; Based on the operation and maintenance task allocation instructions, the technician location information and spare parts inventory location information are obtained, and a personnel dispatch plan and material allocation path are generated through real-time path planning; Record the execution results of the collaborative scheduling decision and calculate the actual execution deviation, generate a historical decision data set based on the execution results and the actual execution deviation, and use the historical decision data set to adjust the prediction parameters and scheduling parameters through an incremental learning model.

[0025] Specifically, through data fusion and two-way feedback prediction, this method greatly improves the accuracy of power generation forecasts, providing a reliable basis for power plants to participate in the power market and accept grid dispatch. Its core collaborative dispatch mechanism realizes the transformation from the previous independent decision-making of each subsystem to global optimization decision-making, effectively balancing the relationship between short-term power generation benefits and long-term equipment health and grid stability, and avoiding the overall losses that may be caused by local optimal decisions. The real-time optimization and dispatch of operation and maintenance resources greatly shortens the fault response and processing time and improves the availability of equipment. Ultimately, the adaptive learning closed loop ensures that the system can continuously adapt to changes in the external environment and internal state, so that the operation and maintenance dispatch strategy of the power station always remains in a near-optimal state, thereby fundamentally enhancing the competitiveness and sustainable operation capabilities of the new energy power station.

[0026] Optionally, generating a unified data matrix includes: Obtain real-time power generation data from new energy power plants, equipment sensor data, weather forecast data, grid dispatch instructions, and power market price information to obtain a basic data set; extracting timestamp and geographic coordinate information from the basic dataset; Synchronizing the data in the basic dataset to a unified time step based on the timestamp, and mapping the data in the basic dataset to a unified spatial reference based on the geographic coordinate information to obtain a processed basic dataset; The processed basic data sets are structured and integrated to generate a unified data matrix.

[0027] Specifically, the process of generating a unified data matrix aims to standardize and integrate data related to new energy power plants from diverse sources and formats, providing a high-quality data foundation for subsequent forecasting and scheduling. This process begins with data collection. The system acquires raw information from multiple sources, comprising the base dataset. This information includes, but is not limited to, real-time power generation data recording the plant's output, sensor data reflecting equipment operating status, weather forecast data predicting future weather conditions, dispatch instructions from the power grid, and electricity market price information reflecting market value. After data collection, the system performs in-depth analysis of the base dataset to extract the core spatiotemporal attributes of each data entry. Specifically, the system identifies and extracts the embedded timestamp information and geographic coordinates that identify its physical source within each data record. Next, based on these spatiotemporal attributes, it performs the crucial spatiotemporal alignment process. In the temporal dimension, the system synchronizes data collected at different frequencies within the base dataset using a predefined, unified time step, such as 15 minutes. For sensor data with a collection frequency higher than the benchmark, such as second-level sensor data, it can be aggregated by the mean or eigenvalue within the time window; for data with a frequency lower than the benchmark, such as hourly weather forecasts, an interpolation algorithm can be used to generate data points at the corresponding time step, thereby ensuring that all data are aligned in the time series. In the spatial dimension, the system maps all data to a unified spatial benchmark based on geographic coordinate information. For example, the sensor data of a specific wind turbine or photovoltaic array is associated with its unique device ID and on-site coordinates to ensure the uniqueness and comparability of the data in physical space. After spatiotemporal alignment, the processed basic data set is obtained. The last step is structured integration. The system organizes all data items in the processed basic data set in a unified format to generate a multi-dimensional unified data matrix. The matrix can be represented as follows: , in, Represents the final unified data matrix. Represents a data point in the matrix. Index i represents a spatial reference identifier, such as a specific power generation equipment number, which is obtained by parsing geographic coordinate information and mapping it to the equipment topology of the power station. Index j represents a unified time step sequence number, which is obtained after synchronization with the original timestamp. Index k represents the feature type of the data, such as active power, equipment temperature, wind speed, electricity price, etc., which is determined by the nature of the original data source. Each element That is, the quantized value of the kth feature of the device numbered i at time j.

[0028] Optionally, the method further includes: Extracting temperature parameters and vibration parameters from the device sensor data; Analyzing the temperature parameter and the vibration parameter using a preset health status threshold to generate a device health status; Based on the health status of the equipment and the time change rates of the temperature parameters and the vibration parameters, a failure risk coefficient is calculated as the equipment maintenance urgency.

[0029] Specifically, the method for calculating equipment maintenance urgency in this invention aims to transform raw equipment sensor data into key indicators guiding operational and maintenance decisions. This process begins with the targeted extraction of core parameters closely related to the mechanical and electrical status of the equipment, namely temperature and vibration parameters, from a unified data matrix. These parameters are direct physical quantities that reflect the health of rotating equipment, such as generators or gearboxes. The system then compares and analyzes the real-time temperature and vibration parameters with preset health thresholds to generate an equipment health assessment. These preset health thresholds are multi-level thresholds determined based on equipment design specifications, industry standards, and statistical analysis of long-term historical operating data, such as normal operating range, warning range, and danger range. By determining the range within which the current parameter value falls, the system can qualitatively or quantitatively classify the equipment health status, for example, outputting a "healthy," "concern," or "alarm" level. To achieve a quantified and dynamic risk assessment from a static assessment, the system not only considers the current equipment health status but also incorporates the time rate of change of the parameters as a key consideration. The system calculates the rate of change of the temperature and vibration parameters over consecutive time steps. Finally, the failure risk factor is calculated based on the equipment health status and the changing trends of key parameters, and is used as the equipment maintenance urgency. This calculation process can be expressed as the following formula: , in, The final generated equipment maintenance urgency is a quantitative value used to prioritize operation and maintenance tasks. It is a discrete score obtained by mapping the device health status ratings such as healthy, concerned, and alarm. The score is directly assigned by the system based on the severity of the health status. and Represents the time rate of change of temperature parameters and vibration parameters respectively, which is obtained by calculating the difference of parameter values ​​in adjacent time steps. Function It is a nonlinear mapping function used to amplify the impact of rapidly changing parameter trends on risk. That is, the faster the parameter changes, the larger the function output value. , , These are preset weight coefficients, which respectively define the importance of the current status of the equipment, temperature change trend, and vibration change trend in the comprehensive risk assessment. These weight coefficients are obtained based on expert experience and historical failure data training.

[0030] Optionally, generating a revised power generation plan includes: Extracting weather trends and historical power generation patterns from the unified data matrix, and performing long-term forecasts to generate long-term forecast results; Extracting real-time device status data from the unified data matrix, performing short-term prediction, and generating short-term prediction results; The prediction error between the long-term prediction result and the short-term prediction result is calculated, and the long-term prediction result is back-propagated and calibrated using the prediction error to generate a revised power generation plan.

[0031] Specifically, the system first performs a long-term forecast. From a previously generated unified data matrix, the system extracts key features that influence power generation capacity over a longer period of time, primarily weather trends and historical power generation patterns. Weather trends include weather forecasts for wind speed, wind direction, and solar irradiance over the next several hours to days, while historical power generation patterns represent the plant's past actual power generation performance under similar meteorological and seasonal conditions. Based on this data, the system employs time series analysis or machine learning models to generate a preliminary long-term forecast, which depicts the plant's ideal power generation capacity curve for the next dispatch cycle. Simultaneously, the system performs a parallel short-term forecast to capture the plant's actual power generation potential in the present and immediate future. The system extracts real-time equipment status data from the unified data matrix to generate a short-term forecast representing the plant's maximum potential output within a very short time window. The key to bidirectional feedback forecasting lies in dynamically calibrating the long-term forecast using the short-term forecast. The system calculates the prediction error between the long-term and short-term forecasts at the current point in time. This error essentially reflects deviations in power generation capacity caused by factors such as real-time equipment status changes that are not fully captured by the long-term model. The system then corrects the long-term forecast results through a back-propagation calibration mechanism, rather than simply replacing the forecast. This calibration can be understood as a bias adjustment, whose impact decays over time. This correction process can be expressed as the following formula: , in, is the revised power generation plan value at time t in the future. is the value of the original long-term forecast result at time t. is the short-term forecast result at the current moment, and is the long-term forecast result at the current moment The difference between the two is the prediction error at the current moment. is a weight decay coefficient that changes with time. Its value is close to 1 at the current moment and gradually decreases to 0 as t increases. This ensures that the calibration effect is most significant in the near future and smoothly transitions to the original long-term forecast in the long term, avoiding the excessive impact of instantaneous disturbances on the entire planning cycle. The sequence is a revised power generation plan that integrates macro-meteorological trends and the real-time status of micro-equipment.

[0032] Optionally, generating a short-term prediction result includes: extracting real-time device status data from the unified data matrix; Using the equipment maintenance urgency, modifying the real-time equipment status data; Perform short-term prediction based on the corrected real-time equipment status data to generate short-term prediction results.

[0033] Specifically, the method for generating short-term prediction results in the present invention is to deeply optimize the assessment of immediate power generation capacity within a two-way feedback prediction framework. The process first extracts real-time equipment status data directly related to the current power generation capacity from a unified data matrix. These data include but are not limited to high-frequency updated operating parameters such as the voltage and current on the DC side of the inverter, the pitch angle of the wind turbine blades, and the speed of the generator. After obtaining the original real-time equipment status data, the system does not use it directly for prediction, but introduces the previously calculated equipment maintenance urgency to dynamically correct these original data. The core idea of ​​this step is that the health status of the equipment will directly limit its theoretical maximum output, so this limitation must be quantified and reflected in the prediction input. The system adopts a derating method to convert the equipment maintenance urgency into an adjustment coefficient for the key performance parameter to generate the corrected real-time equipment status data. For example, for a key performance parameter, its correction process can be expressed as the following formula: , in, It is the corrected real-time device status parameter, such as the corrected maximum available power or conversion efficiency. It is the original real-time equipment status parameter extracted directly from the unified data matrix. is the equipment maintenance urgency calculated previously, which represents the risk of equipment failure. It is a preset upper limit of urgency, used to normalize the urgency of equipment maintenance. Its value is set according to the equipment type and risk management strategy. This is an adjustable sensitivity coefficient, ranging from 0 to 1, that controls the degree to which equipment maintenance urgency affects equipment performance parameter derating. This coefficient is calibrated through historical data analysis and expert experience. This correction process ensures that when equipment maintenance urgency is high, the corresponding performance parameters are appropriately adjusted downward to reflect its true, limited power generation capacity. Finally, the system uses this corrected real-time equipment status data as input to the short-term forecasting model, ultimately generating short-term forecasts that are closer to actual operating limits.

[0034] Optionally, generating a coordinated scheduling decision including a power plant output plan, energy storage charging and discharging strategies, and operation and maintenance task allocation instructions includes: Obtaining the maintenance urgency and energy storage charge status of the equipment; Construct a multi-objective function with power generation efficiency, equipment life, grid demand and economy as optimization targets; Using the revised power generation plan, the equipment maintenance urgency, the energy storage charge state, and the power market price information as input variables of the multi-objective function; The multi-objective function is solved to obtain an optimal balance solution, and the optimal balance solution is parsed into a power plant output plan, an energy storage charging and discharging strategy, and an operation and maintenance task allocation instruction to obtain a collaborative scheduling decision.

[0035] Specifically, the present invention generates collaborative scheduling decisions that include power plant output plans, energy storage charging and discharging strategies, and operation and maintenance task allocation instructions, which are achieved through a multi-objective dynamic trade-off algorithm that aims to find the optimal balance between power generation revenue, equipment health, grid stability, and economic benefits. The process first obtains all the input information required for decision-making, including the previously generated revised power generation plan, the dynamically calculated equipment maintenance urgency, the real-time monitored energy storage charge state, and the electricity price information obtained from the power market. Subsequently, the system constructs a multi-objective function with power generation efficiency, equipment life, grid demand, and economy as optimization objectives. This function is not a single mathematical expression, but an optimization problem framework that aims to simultaneously optimize or satisfy a set of objective functions, which can be expressed as seeking the optimal decision variable set. To optimize the target vector : , In this framework, is a multi-objective optimization function, It is a set of decision variables, mainly including the power plant output plan for each future scheduling time step t , energy storage charging and discharging strategy , positive value is discharge, negative value is charge, and operation and maintenance task allocation instructions A Boolean value or priority index that determines whether maintenance should be performed during this period. is a vector containing the four objective functions. Represents the power generation efficiency target, which aims to maximize the ratio of actual power generation to the theoretical maximum power generation based on the revised power generation plan. Represents the equipment life target, which aims to minimize equipment loss caused by high load operation. This function converts a higher equipment maintenance urgency E into a higher output higher penalties. It represents the grid demand target and aims to minimize the deviation between the actual power plant output and the grid dispatch instructions. It represents the economic goal, which is to maximize the revenue from electricity sales minus the operation and maintenance costs and energy storage loss costs. The revenue from electricity sales is directly related to the electricity price information in the power market.

[0036] During the solution process, the revised power generation plan, equipment maintenance urgency, energy storage charge state, and power market price information are input into the multi-objective function as known parameters or constraints. For example, must be less than or equal to the upper limit given by the revised power generation plan; the charging and discharging behavior of the energy storage is constrained by its current energy storage charge state and maximum / minimum charge state. The system solves the multi-objective function through a multi-objective dynamic trade-off algorithm, such as an optimization algorithm with adaptive weights, to find an optimal balance solution that can best balance the above four conflicting objectives. This solution is essentially a set of optimal value sequences for the decision variable X throughout the entire scheduling cycle. Finally, the system analyzes this optimal balance solution and The sequence is converted into a specific power station output plan, The sequence is converted into energy storage charging and discharging strategy, The sequence is converted into operation and maintenance task allocation instructions, thus forming a unified, internally coordinated collaborative scheduling decision.

[0037] Optionally, generating a personnel dispatch plan and a material allocation path includes: Parsing the operation and maintenance task assignment instructions to determine the required technician skills and spare parts types; Obtain technician location information and spare parts inventory location information; Combine the technician location information and the technician skills to screen out candidate technicians, and combine the spare parts inventory location information and the spare parts type to screen out candidate spare parts libraries; Combine candidate technicians with candidate spare parts libraries, perform path optimization calculations, select combinations based on total response time, and generate personnel dispatch plans and material allocation paths.

[0038] Specifically, by parsing the operation and maintenance task allocation instructions, the specific requirements of each operation and maintenance task can be accurately identified, including the skill level or professional qualifications of the technicians required to perform the task, as well as the type and quantity of spare parts required to complete the repair or replacement. After clarifying the task requirements, the system enters the resource matching stage. It first obtains the current location information of all available technicians and the skill information pre-entered into their files in real time. At the same time, the system will also query the inventory location information of each spare parts, including real-time inventory data of the central warehouse and distributed site spare parts warehouses. Based on the skill requirements required for the task, the system screens out candidate technicians with corresponding qualifications from all technicians. Similarly, based on the required spare parts type, the system screens out candidate spare parts warehouses that have the spare parts and have sufficient inventory.

[0039] Next, the system enters the path optimization calculation phase, aiming to find the resource combination with the shortest response time. The system combines each selected candidate technician with each candidate spare parts library using a Cartesian product, generating multiple potential scheduling scenarios for "personnel-materials-task point." For each scenario, the system calculates its total response time. Calculating total response time is a multi-segment path planning problem, expressed as the following formula: , in, Represents the total response time of the combined solution. It is the shortest travel time between locations calculated by a real-time path planning algorithm that takes into account factors such as real-time traffic conditions and traversable areas determined by geo-fencing technology. is the current location information of the candidate technician, is the location information of the candidate spare parts library, is the location of the equipment to be serviced. The first term in the formula calculates the time it takes a technician to travel to the spare parts depot, and the second term calculates the time it takes to travel from the depot to the service point after carrying the spare parts. is the estimated time required to prepare and receive the spare part. This is a constant based on historical data or preset standards. If the technician can go directly to the task point, for example, the task does not require the spare part or the spare part is already on site, the route planning is simplified to go from the technician's location to the task point. The system will perform this calculation for all possible combinations and select The combination with the smallest value. Ultimately, based on the principle of shortest total response time, the system determines the optimal combination. This solution is then parsed into a specific personnel dispatch plan, specifically which technician is assigned to perform the task, and a detailed material allocation path, specifically which spare parts warehouse the technician should pick up the parts from and then follow the planned optimal route to the task site.

[0040] Optionally, adjusting the prediction parameters and scheduling parameters includes: Recording the execution results of the collaborative scheduling decision and calculating the actual execution deviation, and generating a historical decision data set based on the execution results and the actual execution deviation; Inputting the historical decision dataset into the incremental learning model for training to identify systematic deviation patterns; generating a bias adjustment based on the systematic bias pattern; The deviation adjustment amount is applied to update the prediction parameters and the scheduling parameters.

[0041] Specifically, such as Figure 2 As shown, the process of adjusting the prediction parameters and scheduling parameters in the present invention constructs a closed-loop adaptive learning mechanism, which enables the system to continuously optimize itself from historical experience. The first step of this process is data accumulation and deviation quantification. The system will record in detail the complete execution results of each collaborative scheduling decision, such as the actual output curve of the power station, the actual charging and discharging power of the energy storage, and the actual completion of the operation and maintenance tasks. Subsequently, the system compares these execution results with the original collaborative scheduling decision item by item, and accurately calculates the actual execution deviation, such as the deviation of the power generation plan, the execution error of the energy storage strategy, etc. Based on the original decision, execution results and calculated deviations, the system constructs and continuously expands a structured historical decision data set.

[0042] Next, the system uses the historical decision data set to train an incremental learning model. The characteristic of the incremental learning model is that it can iteratively update the original model when new data arrives, without the need to retrain all historical data, which ensures the efficiency and real-time nature of the learning process. The goal of training is to allow the model to autonomously learn and identify systematic deviation patterns by analyzing historical decision data sets. Systematic deviation patterns refer to regular prediction or decision errors that recur under specific conditions. For example, the model may find that under a certain weather forecast pattern, power generation forecasts are always systematically high. Once the incremental learning model identifies a specific systematic deviation pattern, it can generate a corresponding deviation adjustment based on the pattern. This deviation adjustment can be understood as the model's predicted value of the deviation that may occur in the future under similar conditions. The generation process can be expressed by the following formula: , in, Represents the deviation adjustment generated for a specific forecast parameter or scheduling parameter. It is a trained incremental learning model. The model represents a contextual feature vector describing a future decision moment. This vector includes expected weather conditions, equipment status, electricity prices, and other information at that time. This information is formatted in the same way as the context recorded in the historical decision dataset. The model takes this future context as input and outputs a quantitative prediction of the systematic deviation in that scenario, known as the deviation adjustment. The final step is to apply this deviation adjustment to update the system's core parameters. The system applies the calculated deviation adjustment to the corresponding forecast and scheduling parameters. For example, if the model predicts a systematic negative deviation in power generation for the next period, the system applies this negative deviation adjustment to the relevant parameters of the forecast model to improve the accuracy of future forecasts. Similarly, if a dispatch weight setting is found to result in consistently suboptimal results, the weight will be adjusted accordingly. Through this continuous cycle of "recording, analyzing, adjusting, and applying," the system's forecasting and scheduling capabilities are continuously iterated and refined.

[0043] Optionally, the method further includes: Based on the revised power generation plan, determining whether a sudden drop in future power generation will occur; If so, trigger an emergency response, calculate the predicted power gap based on the revised power generation plan, and update the energy storage charging and discharging strategy according to the power gap; The operation and maintenance task allocation instructions are synchronously adjusted, and adaptive scheduling instructions are generated in combination with the updated energy storage charging and discharging strategy.

[0044] Specifically, the emergency response method for handling sudden drops in power generation in the present invention is a set of proactive risk management mechanisms based on accurate predictions. The triggering condition for this process is the system's real-time monitoring of future power generation. The system will continuously analyze the revised power generation plan generated by the two-way feedback prediction mechanism and calculate its first-order difference between adjacent time steps, that is, the rate of change of power generation. When the system detects that the negative value of this rate of change exceeds a preset sudden drop threshold, for example, when the power generation is predicted to drop significantly in a short period of time, the emergency response process will be automatically triggered. Once the emergency response is triggered, the system will first accurately calculate the expected power gap based on the current revised power generation plan. The power gap is calculated by comparing the revised power generation plan sequence within a period of time after the sudden drop (for example, the next 1 hour) with the power generation plan or grid dispatch instruction in the stable state before the sudden drop. The sum of the differences is the predicted power gap. The calculation can be expressed by the following formula: , in, Represents the predicted total electricity gap. is the target output at time t, which can be the stable output value before the sudden drop or the dispatch instruction value issued by the power grid. is the revised power generation plan value at time t after the emergency response is triggered. The pre-set emergency response time window. After quantifying the power gap, the system immediately updates the existing energy storage charging and discharging strategy based on the gap. The system replans the energy storage system's discharge plan to discharge according to the optimal power curve during the power gap period, aiming to maximize the reduction in power generation and smooth fluctuations in the power plant's total output. This update process considers constraints such as the energy storage's current state of charge, maximum discharge power, and remaining available power to ensure the strategy's feasibility. Simultaneously, the emergency response mechanism also adjusts the operation and maintenance task allocation instructions. The system analyzes the root cause of the sudden drop in power generation. If the drop in the revised power generation plan is due to a predicted failure or a sharp deterioration in the performance of a specific piece of equipment, this information is incorporated into the generation of the revised power generation plan. The system immediately increases the maintenance task priority for the relevant equipment and may generate new, urgent maintenance task allocation instructions. This step ensures that maintenance resources are quickly directed to the root cause of the problem. Finally, the system integrates the updated energy storage charging and discharging strategy with the adjusted maintenance task allocation instructions to generate a new set of adaptive scheduling instructions tailored to the current emergency situation. This set of instructions will replace the corresponding parts of the original coordinated scheduling decision-making and be sent down to the power station's execution units, including the energy storage control system and operation and maintenance management platform, to achieve a rapid and coordinated response to emergencies.

[0045] Based on the same inventive concept, Figure 3 As shown, the present invention also provides a new energy power station intelligent operation and maintenance scheduling and resource optimization system, the system comprising: The data fusion module is used to obtain real-time power generation data of new energy power plants, equipment sensor data, weather forecast data, grid dispatch instructions, and power market electricity price information, and generate a unified data matrix through spatiotemporal alignment processing; a collaborative forecasting module, configured to perform bidirectional feedback forecasting based on the unified data matrix, the bidirectional feedback forecasting including generating a long-term forecast result and a short-term forecast result, and using the short-term forecast result to perform error correction on the long-term forecast result to generate a revised power generation plan; An intelligent scheduling module is used to obtain the equipment maintenance urgency and energy storage charge state, and combine the revised power generation plan with the electricity market price information to generate a coordinated scheduling decision including the power station output plan, energy storage charging and discharging strategy, and operation and maintenance task allocation instructions through a multi-objective dynamic trade-off algorithm; A resource optimization module is used to obtain technician location information and spare parts inventory location information based on the operation and maintenance task allocation instructions, and generate personnel dispatch plans and material allocation paths through real-time path planning; An adaptive learning module is used to record the execution results of the collaborative scheduling decision and calculate the actual execution deviation, generate a historical decision data set based on the execution results and the actual execution deviation, and use the historical decision data set to adjust the prediction parameters and scheduling parameters through an incremental learning model.

[0046] To verify the feasibility of this invention, it was applied to a new energy demonstration power station. This power station includes a large-scale photovoltaic array, wind turbines, and a supporting energy storage system. The goal was to optimize power generation planning, improve operational efficiency, and maximize market returns. After deployment, the system was validated over several months using real-time data collection, scheduling decision generation, and closed-loop optimization.

[0047] In this embodiment, the data fusion module of the present invention first processes the multi-source heterogeneous data of the power station. The system obtains the real-time power generation data of the photovoltaic array, the equipment sensor data of the wind turbine such as the W11 wind turbine including temperature and vibration parameters, the regional minute-level weather forecast data, the dispatch instructions issued by the power grid, and the real-time electricity price information of the provincial power market. Through spatiotemporal alignment processing, the system synchronizes these data of different frequencies and formats to a unified 15-minute time step, and maps them to a unified spatial reference such as a specific device ID, generating a structured unified data matrix, which can be expressed as , providing a high-quality data foundation for subsequent accurate prediction and scheduling.

[0048] At 10:15 AM on August 5, 2024, the system detected from the unified data matrix that the vibration parameters of wind turbine W11's gearbox were showing a slight upward trend. Based on the preset health threshold, the system adjusted its health status from "Healthy" to "Concern." The system also calculated the equipment maintenance urgency based on the time-varying rate of change of the vibration parameters using the following formula: The calculation shows that the equipment maintenance urgency value has increased from the daily 15 to 45 out of 100, indicating that there is a potential risk of failure.

[0049] Subsequently, the collaborative forecast module uses this urgency to correct the real-time status data of wind turbine W11 when generating the short-term forecast results. , the system derates its theoretical maximum available power parameters. This revised state data is input into the short-term prediction model to generate a short-term prediction result that reflects its actual power generation capacity. At the same time, the long-term prediction model gives a higher power generation forecast based on the weather forecast (stable wind speed). The two-way feedback prediction mechanism is activated, and through the formula Using the lowered short-term forecast, the team back-propagated the long-term forecast to calibrate it, generating a more accurate revised power generation plan. This plan shows that over the next three hours, the total power plant output will be approximately 5MW lower than the original ideal forecast.

[0050] In the intelligent dispatching phase, the intelligent dispatching module obtained the revised power generation plan, the high maintenance urgency of wind turbine W11, the 70% charge state of the energy storage system, and the electricity price information during the peak electricity price period in the market. , making dynamic trade-offs. To balance economic efficiency (high electricity prices) and equipment lifespan (high urgency), the algorithm generated a coordinated dispatch decision: 1. The power plant's output plan will be implemented according to the revised power generation plan for the next hour; 2. The energy storage system will discharge 5MWh within one hour to compensate for the potential output drop of wind turbine W11, ensuring that grid dispatch instructions are met and maximum electricity sales revenue is achieved; 3. An operation and maintenance task is assigned to wind turbine W11, scheduling it for maintenance two hours later during a period of lower wind speeds.

[0051] Based on the operation and maintenance task allocation instruction, the resource optimization module is activated. The instruction analysis determines that a mechanic with a second-level qualification and a specific type of gearbox lubricant are required. The system obtains the technician Zhang Gong with a second-level qualification and is located in the main control building, and Li Gong with a second-level qualification and is on patrol in Area F, as well as the location information of Zhang Gong and Li Gong, as well as the inventory location information of the central spare parts warehouse A and the No. 3 distributed spare parts warehouse B. Both have the required spare parts. Through path optimization calculation The system calculated that the total response time for "Engineer Zhang departing from the main control building, going to spare parts warehouse A to collect supplies, and then to wind turbine W11" was 45 minutes, while other combinations took over an hour. Therefore, the system generated a personnel dispatch plan, assigned Engineer Zhang to the task, and planned the optimal material allocation route.

[0052] In an emergency, on September 2, 2024, a cloud moved rapidly and blocked the photovoltaic area. The power generation plan was revised to determine that the power generation would drop by 30MW in the next 15 minutes. The system immediately triggered an emergency response and used the formula The predicted power gap was calculated, and the energy storage charging and discharging strategy was updated based on this gap. The energy storage system was instructed to discharge at maximum power for emergency support, generating adaptive scheduling instructions and successfully avoiding drastic fluctuations in power output to the grid.

[0053] After a quarter of operation, the adaptive learning module recorded the deviations between the execution results and actual execution of all decisions and generated a historical decision data set. Through incremental learning model training, the system found that in cloudy to sunny weather patterns, its prediction of the PV output ramp rate is always systematically slow. Based on this deviation pattern, the model uses the formula A bias adjustment was generated and applied to update the relevant weight parameters in the forecast model. This adjustment improved the system's forecast accuracy for that particular weather condition from 88% to 95%.

[0054] Table 1 New energy power station coordinated dispatch decision data table Table 2 Operation and maintenance resource optimization scheduling data table Table 3 System adaptive learning adjustment data table Tables 1 through 3 above record actual application data of the present invention in a new energy demonstration power station, detailing the system's performance in collaborative scheduling, resource optimization, and adaptive learning. These data clearly demonstrate the advanced nature and practical value of the present invention.

[0055] As shown in Table 1, on August 5, 2024, the system successfully addressed the potential failure risk of wind turbine W11 by revising the power generation plan and implementing energy storage scheduling, ensuring equipment safety while maintaining stable power output. Furthermore, on September 2, faced with a sudden drop in power generation, the system implemented an emergency response mechanism and utilized energy storage to quickly compensate for the power shortfall, resulting in only minor fluctuations in actual output, demonstrating its exceptional risk mitigation capabilities.

[0056] Table 2 shows how the system implements quantitative optimization in resource allocation for O&M tasks. By accurately calculating the response times for different personnel-material combinations, the system selects the optimal solution, reducing response time from over an hour to 45 minutes, significantly improving O&M efficiency.

[0057] Table 3 demonstrates the system's powerful self-learning and self-evolution capabilities. By analyzing and learning from historical deviations, the system automatically adjusts its internal model parameters, continuously improving forecast accuracy and the economic efficiency of scheduling strategies. For example, the forecast accuracy for specific weather patterns increased by 7 percentage points, directly translating into better power generation plans and higher market returns. These data demonstrate the significant technical effectiveness of this invention in improving the intelligence level, operational efficiency, and economic benefits of new energy power plants.

[0058] It should be noted that the formulas appearing above can translate physical quantities of different properties into unitless standard values ​​or superimposable parameters of the same dimension through the principle of dimensional consistency and mathematical standardization means (such as normalization, dimensionless parameter conversion or unit system unification), thereby eliminating the interference of different dimensions on the operation logic, so that the formulas have mathematical operation rationality and objective law adaptability while retaining the distribution characteristics of the original data. It is a conventional technical means and will not be elaborated here. The electrical connection between the above-mentioned units does not necessarily mean a direct connection of the circuit. The indirect connection method can be applied to the embodiments of the present invention as long as the purpose of the present invention is achieved. The above is only an exemplary embodiment of the present invention and the scope of the present invention cannot be limited thereto.

[0059] That is, any equivalent changes and modifications made according to the teachings of the present invention are still within the scope of the present invention. Those skilled in the art will readily conceive of other embodiments of the present invention after considering the disclosure of the specification and practical truths. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary technical means in the art not described herein.

Claims

1. A method for intelligent operation and maintenance scheduling and resource optimization of a new energy power station, characterized in that: The method comprises: Acquire real-time power generation data from new energy power stations, equipment sensor data, weather forecast data, grid dispatch instructions, and power market price information, and generate a unified data matrix through spatiotemporal alignment processing; Based on the unified data matrix, performing bidirectional feedback forecasting, the bidirectional feedback forecasting including generating a long-term forecast result and a short-term forecast result, and using the short-term forecast result to perform error correction on the long-term forecast result to generate a revised power generation plan; Obtaining the equipment maintenance urgency and energy storage charge state, and combining the revised power generation plan with the electricity market price information, generating a collaborative scheduling decision including the power station output plan, energy storage charging and discharging strategy, and operation and maintenance task allocation instructions through a multi-objective dynamic trade-off algorithm; Based on the operation and maintenance task allocation instructions, the technician location information and spare parts inventory location information are obtained, and a personnel dispatch plan and material allocation path are generated through real-time path planning; Record the execution results of the collaborative scheduling decision and calculate the actual execution deviation, generate a historical decision data set based on the execution results and the actual execution deviation, and use the historical decision data set to adjust the prediction parameters and scheduling parameters through an incremental learning model.

2. A method for intelligent operation and maintenance scheduling and resource optimization of a new energy power station according to claim 1, characterized in that: Generating a unified data matrix comprises: Obtain real-time power generation data from new energy power plants, equipment sensor data, weather forecast data, grid dispatch instructions, and power market price information to obtain a basic data set; extracting timestamp and geographic coordinate information from the basic dataset; Synchronizing the data in the basic dataset to a unified time step based on the timestamp, and mapping the data in the basic dataset to a unified spatial reference based on the geographic coordinate information to obtain a processed basic dataset; The processed basic data sets are structured and integrated to generate a unified data matrix.

3. The method for intelligent operation and maintenance scheduling and resource optimization of a new energy power station according to claim 1 is characterized in that: The method further comprises: Extracting temperature parameters and vibration parameters from the device sensor data; Analyzing the temperature parameter and the vibration parameter using a preset health status threshold to generate a device health status; Based on the health status of the equipment and the time change rates of the temperature parameters and the vibration parameters, a failure risk coefficient is calculated as the equipment maintenance urgency.

4. A method for intelligent operation and maintenance scheduling and resource optimization of a new energy power station according to claim 3, characterized in that: Generating a modified power generation plan includes: Extracting weather trends and historical power generation patterns from the unified data matrix, and performing long-term forecasts to generate long-term forecast results; Extracting real-time device status data from the unified data matrix, performing short-term prediction, and generating short-term prediction results; The prediction error between the long-term prediction result and the short-term prediction result is calculated, and the long-term prediction result is back-propagated and calibrated using the prediction error to generate a revised power generation plan.

5. A method for intelligent operation and maintenance scheduling and resource optimization of a new energy power station according to claim 4, characterized in that: Generating a short-term prediction result includes: extracting real-time device status data from the unified data matrix; Using the equipment maintenance urgency, modifying the real-time equipment status data; Perform short-term prediction based on the corrected real-time equipment status data to generate short-term prediction results.

6. A method for intelligent operation, maintenance, scheduling and resource optimization of a new energy power station according to claim 4, characterized in that: The generation of a coordinated scheduling decision including a power plant output plan, energy storage charging and discharging strategy, and operation and maintenance task allocation instructions includes: Obtaining the maintenance urgency and energy storage charge status of the equipment; Construct a multi-objective function with power generation efficiency, equipment life, grid demand and economy as optimization targets; Using the revised power generation plan, the equipment maintenance urgency, the energy storage charge state, and the power market price information as input variables of the multi-objective function; The multi-objective function is solved to obtain an optimal balance solution, and the optimal balance solution is parsed into a power plant output plan, an energy storage charging and discharging strategy, and an operation and maintenance task allocation instruction to obtain a collaborative scheduling decision.

7. A method for intelligent operation, maintenance, scheduling and resource optimization of a new energy power station according to claim 6, characterized in that: The generation of the personnel dispatch plan and material allocation path includes: Parsing the operation and maintenance task assignment instructions to determine the required technician skills and spare parts types; Obtain technician location information and spare parts inventory location information; Combine the technician location information and the technician skills to screen out candidate technicians, and combine the spare parts inventory location information and the spare parts type to screen out candidate spare parts libraries; Combine candidate technicians with candidate spare parts libraries, perform path optimization calculations, select combinations based on total response time, and generate personnel dispatch plans and material allocation paths.

8. The method for intelligent operation, maintenance, scheduling and resource optimization of a new energy power station according to claim 6, characterized in that: The adjustment of prediction parameters and scheduling parameters includes: Recording the execution results of the collaborative scheduling decision and calculating the actual execution deviation, and generating a historical decision data set based on the execution results and the actual execution deviation; Inputting the historical decision dataset into an incremental learning model for training to identify systematic deviation patterns; generating a bias adjustment based on the systematic bias pattern; The deviation adjustment amount is applied to update the prediction parameters and the scheduling parameters.

9. A method for intelligent operation, maintenance, scheduling and resource optimization of a new energy power station according to claim 6, characterized in that: The method further comprises: Based on the revised power generation plan, determining whether a sudden drop in future power generation will occur; If so, trigger an emergency response, calculate the predicted power gap based on the revised power generation plan, and update the energy storage charging and discharging strategy according to the power gap; The operation and maintenance task allocation instructions are synchronously adjusted, and adaptive scheduling instructions are generated in combination with the updated energy storage charging and discharging strategy.

10. A new energy power station intelligent operation and maintenance scheduling and resource optimization system, applied to the new energy power station intelligent operation and maintenance scheduling and resource optimization method according to any one of claims 1 to 9, characterized in that: The system comprises: The data fusion module is used to obtain real-time power generation data of new energy power plants, equipment sensor data, weather forecast data, grid dispatch instructions, and power market electricity price information, and generate a unified data matrix through spatiotemporal alignment processing; a collaborative forecasting module, configured to perform bidirectional feedback forecasting based on the unified data matrix, the bidirectional feedback forecasting including generating a long-term forecast result and a short-term forecast result, and using the short-term forecast result to perform error correction on the long-term forecast result to generate a revised power generation plan; An intelligent scheduling module is used to obtain the equipment maintenance urgency and energy storage charge state, and combine the revised power generation plan with the electricity market price information to generate a coordinated scheduling decision including the power station output plan, energy storage charging and discharging strategy, and operation and maintenance task allocation instructions through a multi-objective dynamic trade-off algorithm; A resource optimization module is used to obtain technician location information and spare parts inventory location information based on the operation and maintenance task allocation instructions, and generate personnel dispatch plans and material allocation paths through real-time path planning; An adaptive learning module is used to record the execution results of the collaborative scheduling decision and calculate the actual execution deviation, generate a historical decision data set based on the execution results and the actual execution deviation, and use the historical decision data set to adjust the prediction parameters and scheduling parameters through an incremental learning model.

Images