Multi-element scheduling pre-decision method, device and equipment for pumped storage power station and medium

Abstract

The invention relates to the technical field of power optimization scheduling, and discloses a multivariate scheduling pre-decision method, device and equipment for a pumped storage power station and a medium. Clustering historical electricity price data in the structured data by using a machine learning clustering algorithm, and identifying a typical day scene; automatically matching a mixed integer linear programming model framework corresponding to the typical day scenes, and outputting to-be-solved parameterized mathematical optimization model instances of which the number is the same as that of the typical day scenes; according to the method, multi-source data collection is achieved by collecting structured power data and unstructured standard file data, a machine learning clustering algorithm is used for replacing a scheduling hypothesis of a traditional fixed mode, a mixed integer linear programming model is used for solving, and the optimized scheduling curve of the pumped storage power station is output. The method is provided for decision makers to call, and provides data support for comprehensive evaluation of the pumped storage power station.

Description

Technical Field

[0001] This invention relates to the field of power optimization dispatching technology, specifically to a multi-factor dispatching pre-decision method, device, equipment, and medium for pumped storage power stations. Background Technology

[0002] Currently, with the large-scale integration of renewable energy sources such as wind power and photovoltaics into the power grid, the demand for peak shaving, frequency regulation, and energy storage in the power system is increasing. Pumped storage power stations, with their advantages of rapid response and strong peak shaving and frequency regulation capabilities, have become a key regulatory resource for the safe and stable operation of the power grid and an ancillary service resource in the diversified electricity market. Existing technologies often focus on a single market or a single indicator, relying on fixed parameter settings, which limits their application in making decisions regarding pumped storage power stations. Summary of the Invention

[0003] This invention provides a multi-dimensional scheduling pre-decision method, apparatus, equipment, and medium for pumped storage power plants, in order to solve the limitations of existing technologies in making decisions for pumped storage power plants.

[0004] In a first aspect, the present invention provides a multi-factor dispatch pre-decision method for pumped storage power stations, the method comprising: Acquire raw structured data related to electricity and unstructured data related to standard documents, and convert the unstructured data into structured data. The raw structured data and the structured data obtained from the conversion of unstructured data together constitute all the structured data. Machine learning clustering algorithms are used to cluster historical electricity price data in all structured data to identify typical daily scenarios, including peak days, average days, and off-peak days. Automatically match the mixed-integer linear programming model framework corresponding to typical daily scenarios, and output the same number of parameterized mathematical optimization model instances to be solved as the typical daily scenarios; Solve the parameterized mathematical optimization model instance using the solver and output the optimized scheduling curve of the pumped storage power station.

[0005] This invention achieves multi-source data acquisition by collecting structured power data and unstructured standard file data. It uses machine learning clustering algorithms to automatically identify typical daily scenarios, replacing the traditional fixed-pattern scheduling assumptions, to accurately match dynamically changing market electricity prices. It uses a mixed-integer linear programming model to solve for and optimize scheduling curves, improving the scientific nature of scheduling decisions. The data is available for decision-makers and provides data support for the comprehensive evaluation of pumped storage power stations, significantly improving its practicality.

[0006] In one alternative implementation, converting unstructured data into structured data includes: Using optical character recognition technology to extract text content from unstructured data; Key numerical parameters are extracted from text content using predefined rules and keyword matching algorithms, and then transformed into structured data. These key numerical parameters include capacity and electricity price.

[0007] This invention combines optical character recognition technology, predefined rules, and keyword matching algorithms to convert unstructured data into structured data, making it suitable for subsequent machine algorithms while ensuring the accuracy of the data conversion.

[0008] In one optional implementation, a machine learning clustering algorithm is used to cluster historical electricity price data from all structured data to identify typical daily scenarios, including: The K-means clustering algorithm is used to perform unsupervised learning on historical electricity price data in structured data. The optimization objective is to minimize the sum of squared errors of all sample points within a cluster, and data with similar electricity price trends are aggregated into different clusters. The center vectors of each cluster are sorted according to the time series to generate the daily electricity price curve; Descriptive labels are automatically matched to the daily electricity price curve to determine typical daily scenarios. The descriptive labels include peak electricity price days, off-peak electricity price days, and low-peak electricity price days.

[0009] This invention automatically identifies typical daily scenarios from historical electricity price data, abandoning the traditional simplified and fixed operational strategy assumptions. It can match the optimal operating mode for pumped storage power stations based on real and dynamically changing market price signals.

[0010] In one optional implementation, a solver is used to solve a parameterized mathematical optimization model instance, outputting an optimized scheduling curve for the pumped storage power station, including: With the objective function of minimizing the overall operating cost of pumped storage power stations, and with hydraulic constraints, physical constraints, electricity market coupling constraints, and relevant document boundary constraints as constraints, a parameterized mathematical optimization model instance is solved to obtain a time series array, which is then used to construct the optimal scheduling curve for pumped storage power stations.

[0011] This invention aims to minimize the overall operating cost of pumped storage power stations. By solving a parameterized mathematical optimization model instance under multiple constraints, it obtains the optimized extraction curve of the pumped storage power station. Compared with traditional calculations based on experience or simple rules, this invention can effectively improve accuracy.

[0012] In one alternative implementation, hydraulic constraints include reservoir water balance equations, physical constraints include physical boundaries of unit operation, and market coupling constraints include ensuring that the sum of power generation, increased reserve capacity, and decreased reserve capacity does not exceed the unit's maximum available output.

[0013] This invention addresses the core pain points in both theory and practice by leveraging the synergistic effects of comprehensive hydraulic constraints, physical constraints, and market-coupled constraints to generate an optimized scheduling curve that conforms to both physical equipment capabilities and market rules.

[0014] In one alternative implementation, after solving the parameterized mathematical optimization model instance using a solver and outputting the optimal scheduling curve for the pumped storage power station, the method further includes: By using power production simulation based on unit combination and economic dispatch, the differences in technical indicators between the baseline scenario and the evaluation scenario are compared. The differences between the optimized scheduling curves and technical indicators are collected to generate a pre-decision report.

[0015] This invention quantifies technical indicators by comparing the differences in technical indicators between benchmark and evaluation scenarios, automatically collects data, and generates a pre-decision report for decision-makers to use, providing data support for the comprehensive evaluation of pumped storage power stations and significantly improving practicality.

[0016] In one alternative implementation, power production simulation based on unit combination and economic dispatch is used to compare the differences in technical indicators between the baseline scenario and the evaluation scenario, including: The scenario without pumped storage power stations is used as the baseline scenario, and the scenario with pumped storage power stations and the same power grid model and boundaries as the baseline scenario is used as the evaluation scenario. Run a production simulation of a benchmark scenario to obtain benchmark technical indicators; Production simulations of the evaluation scenarios are run to obtain evaluation technical indicators; The differences between the technical indicators of the benchmark and evaluation scenarios are obtained by performing item-by-item differential calculations on the benchmark and evaluation technical indicators.

[0017] This invention provides objective and reliable data for power grid dispatch by running production simulations of the benchmark and evaluation scenarios using the same power grid model and boundaries, and performing differential calculations on the obtained evaluation technical indicators to avoid interference from other factors.

[0018] In one alternative implementation, the differences between optimized scheduling curves and technical indicators are aggregated to generate a pre-decision report, including: Load the decision report template, which includes a core control instruction area and a technical benefit analysis area; The optimized scheduling curve is displayed in the core control instruction area, and the expected technical benefits of the scheduling strategy are displayed in the technical benefit analysis area based on the differences in technical indicators, generating a pre-decision report.

[0019] This invention displays the expected technical benefits of optimized scheduling curves and scheduling strategies in the core control instruction area and technical benefit analysis area of ​​the decision report template, respectively. It adopts a visualization combined with a hierarchical design to provide data reference for decision-makers.

[0020] Secondly, the present invention provides a multi-factor dispatch pre-decision device for pumped-storage power stations, the device comprising: The acquisition module is used to acquire raw structured data related to electricity and unstructured data related to standard documents, and convert the unstructured data into structured data. The raw structured data and the structured data obtained from the conversion of unstructured data together constitute all the structured data. The clustering module is used to cluster historical electricity price data in all structured data using machine learning clustering algorithms to identify typical daily scenarios, including peak days, average days, and off-peak days. The matching module is used to automatically match the mixed-integer linear programming model framework corresponding to the typical daily scenario and output the same number of parameterized mathematical optimization model instances to be solved as the typical daily scenario. The solver module is used to solve parameterized mathematical optimization model instances using a solver and output optimized scheduling curves for pumped storage power stations. The production simulation module is used to compare the differences in technical indicators between the baseline scenario and the evaluation scenario by using power production simulation based on unit combination and economic dispatch. The report generation module is used to collect the differences in optimized scheduling curves and technical indicators, and generate a pre-decision report.

[0021] In one optional implementation, the acquisition module includes: The first extraction unit is used to extract text content from unstructured data using optical character recognition technology; The second extraction unit is used to extract key numerical parameters from the text content using predefined rules and keyword matching algorithms, and to convert the key numerical parameters into structured data. The key numerical parameters include capacity electricity price.

[0022] In one alternative implementation, the clustering module includes: The aggregation unit is used to perform unsupervised learning on historical electricity price data in structured data using the K-means clustering algorithm. The optimization objective is to minimize the sum of squared errors of all sample points within a cluster, and to aggregate data with similar electricity price trends into different clusters. The sorting unit is used to sort the center vectors of each cluster according to the time series to generate the daily electricity price curve; The determination unit is used to automatically match descriptive labels to the daily electricity price curve and determine typical daily scenarios. The descriptive labels include peak electricity price days, off-peak electricity price days, and low-peak electricity price days.

[0023] In one alternative implementation, the solver module includes: The solution unit is used to solve a parameterized mathematical optimization model instance with the objective function of minimizing the overall operating cost of the pumped storage power station and with hydraulic constraints, physical constraints, electricity market coupling constraints, and relevant document boundary constraints as constraints. The result is a time series array, which is used to construct the optimal scheduling curve of the pumped storage power station.

[0024] In one alternative implementation, hydraulic constraints include reservoir water balance equations, physical constraints include physical boundaries of unit operation, and market coupling constraints include ensuring that the sum of power generation, increased reserve capacity, and decreased reserve capacity does not exceed the unit's maximum available output.

[0025] In one alternative embodiment, the device further includes: The production simulation module is used to compare the differences in technical indicators between the baseline scenario and the evaluation scenario by using power production simulation based on unit combination and economic dispatch. The report generation module is used to collect the differences in optimized scheduling curves and technical indicators, and generate a pre-decision report.

[0026] In one alternative implementation, the production simulation module includes: The scenario determination unit is used to take a scenario that does not include a pumped storage power station as the baseline scenario and a scenario that includes a pumped storage power station and has the same power grid model and boundaries as the baseline scenario as the evaluation scenario. The first production simulation unit is used to run production simulations of benchmark scenarios to obtain benchmark technical indicators. The second production simulation unit is used to run production simulations of the evaluation scenario to obtain evaluation technical indicators. The differential calculation unit is used to perform differential calculations on the benchmark technical indicators and the evaluation technical indicators one by one to obtain the technical indicator differences between the benchmark scenario and the evaluation scenario.

[0027] In one alternative implementation, the report generation module includes: The loading unit is used to load the decision report template, which includes a core control instruction area and a technical benefit analysis area. The report generation unit is used to display the optimized scheduling curve in the core control instruction area, and based on the differences in technical indicators, to display the expected technical benefits of the scheduling strategy in the technical benefit analysis area, and generate a pre-decision report.

[0028] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the multi-factor dispatch pre-decision method for pumped storage power stations described in the first aspect or any corresponding embodiment thereof.

[0029] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the multi-factor dispatch pre-decision method for pumped storage power stations described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0030] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0031] Figure 1 This is a flowchart illustrating a multi-factor dispatch pre-decision method for pumped storage power plants according to an embodiment of the present invention. Figure 2 This is a schematic diagram illustrating the revenue calculation according to an embodiment of the present invention; Figure 3 This is a structural block diagram of a multi-element dispatch pre-decision device for pumped storage power stations according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0032] 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 only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0034] In related technologies, existing revenue calculations often focus on energy market arbitrage (i.e., pumping during off-peak hours and generating during peak hours) or compensation for single ancillary services (such as frequency regulation and reserve capacity), often neglecting capacity subsidies, electricity pricing documents, renewable energy integration benefits, and the overall reduction in grid operating costs. Meanwhile, the electricity market consists of medium- and long-term contracts, day-ahead spot markets, real-time spot markets, and ancillary service markets. The settlement rules and pricing mechanisms of each market are constantly evolving, making it difficult to accurately reflect the comprehensive value of pumped storage power plants using methods based on a single market or single indicator.

[0035] To address the aforementioned issues, this invention provides a multi-dimensional dispatch pre-decision method for pumped storage power stations. Its core objective is to overcome the limitations of traditional methods for calculating and dispatching pumped storage power station revenue, providing more accurate and comprehensive technical support for investment decisions, planning and design, and operation management. This helps pumped storage power stations maximize their benefits and enhance their value in a diversified electricity market environment. The method organically integrates standard documents, market prices, simulations, and optimized dispatching to achieve end-to-end calculations and dynamically quantify the annual comprehensive revenue of pumped storage power stations in complex electricity market environments.

[0036] The multi-dimensional dispatch pre-decision application for pumped storage power stations is based on a hardware architecture consisting of one or more servers. The hardware architecture includes a central processing unit (CPU), memory (RAM), storage devices, and network interfaces. The CPU is used to execute dispatch algorithms and model calculations. The RAM is used to temporarily store programs and data during processing. The storage devices are used to permanently store historical data, policy text libraries, optimization models, and calculation results. The network interface is used to obtain real-time and historical data from the power grid dispatch automation system (EMS) and the power market trading platform through the Enterprise Service Bus (ESB) or a dedicated network channel. The dispatch algorithms and models are stored in the storage devices as software programs and are loaded into RAM by the CPU for execution.

[0037] According to an embodiment of the present invention, a multi-dispatch pre-decision method for pumped storage power stations is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0038] This embodiment provides a multi-factor dispatch pre-decision method for pumped storage power stations. Figure 1 This is a flowchart of a multi-factor dispatch pre-decision method for pumped storage power stations according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: Step S101: Obtain the original structured data related to electricity and the unstructured data related to standard documents, and convert the unstructured data into structured data. The original structured data and the structured data obtained from the conversion of the unstructured data together constitute all the structured data.

[0039] In this embodiment of the invention, a network interface is used to call a pre-defined interface (such as an API, Application Programming Interface) to automatically obtain structured data such as day-ahead electricity prices and reserve prices from the electricity market trading platform database at a preset time frequency. For example, structured data is automatically obtained at fixed time points to ensure the timeliness and accuracy of the data.

[0040] Understandably, after acquiring structured data, it undergoes cleaning and standardization processes, including validating data formats, identifying and handling outliers, filling in missing values, and standardizing units of measurement. The preprocessed structured data is then stored in a locally deployed time series database (TSDB) for quick querying and retrieval by subsequent optimization models.

[0041] The system automatically scans and monitors a standard document repository to obtain unstructured data. This repository includes a policy document library, which can be a local disk directory or a network shared drive. When a new policy document or a policy document update is detected, the unstructured data is automatically converted into structured data and stored in a parameter database. This automates the entry of policy information, improves the accuracy of the entry, and solves the problems of low efficiency and error-proneness in manually interpreting policy documents.

[0042] The original structured data and the structured data obtained by transforming unstructured data together constitute all structured data, serving as the data foundation.

[0043] Step S102: Use machine learning clustering algorithms to cluster historical electricity price data in all structured data to identify typical daily scenarios.

[0044] In this embodiment of the invention, a machine learning clustering algorithm is invoked to perform clustering processing on all structured data. This algorithm is used to identify typical daily scenarios, including but not limited to peak days, average days, and low-end days.

[0045] Step S103: Automatically match the mixed-integer linear programming model framework corresponding to the typical daily scenario, and output the same number of parameterized mathematical optimization model instances to be solved as the typical daily scenario.

[0046] In this embodiment of the invention, a preset Mixed-Integer Linear Programming (MILP) basic model framework is automatically matched for each typical daily scenario. For example, for scenarios with large peak-valley price differences, a model framework with energy time-shift arbitrage as the main objective is matched; for scenarios with extremely large peak-valley price differences, a model framework with the objective of maximizing energy time-shift arbitrage is matched. For scenarios with high reserve prices, a model framework focusing on optimizing reserve capacity allocation is matched; for scenarios with small price differences but high reserve prices, a model framework with the objective of optimal reserve capacity allocation is matched.

[0047] By combining the well-matched mixed-integer linear programming model framework with the characteristic parameters of a typical daily scenario (such as the electricity price curve and reserve price of the scenario), a number of parameterized mathematical optimization model instances to be solved are generated, which are the same as those of the typical daily scenario, in preparation for the next step of solution calculation.

[0048] Step S104: Use the solver to solve the parameterized mathematical optimization model instance and output the optimized scheduling curve of the pumped storage power station.

[0049] In this embodiment of the invention, a mixed-integer linear programming solver is activated. The solver receives model instances of typical daily scenarios, finds the optimal solution that meets the conditions, and outputs the optimized scheduling curve of the pumped storage power station. The curve includes at least the power generation / pumping power, the increase in reserve capacity, and the decrease in reserve capacity at each time moment.

[0050] The multi-source scheduling pre-decision method for pumped storage power stations provided in this embodiment achieves multi-source data acquisition by collecting structured power data and unstructured standard file data. It uses machine learning clustering algorithms to automatically identify typical daily scenarios, replacing the traditional fixed-pattern scheduling assumptions, to accurately match dynamically changing market electricity prices. It uses a mixed-integer linear programming model to solve for and optimize the scheduling curve, improving the scientific nature of scheduling decisions. This method is available for decision-makers and provides data support for the comprehensive evaluation of pumped storage power stations, significantly improving its practicality.

[0051] This embodiment provides a multi-factor dispatch pre-decision method for pumped storage power stations, the process of which includes the following steps: Step S201: Obtain the original structured data related to electricity and the unstructured data related to standard documents, and convert the unstructured data into structured data.

[0052] Specifically, step S201 above, which converts unstructured data into structured data, includes: Step S2011: Use optical character recognition technology to extract text content from unstructured data.

[0053] Step S2012: Use predefined rules and keyword matching algorithms to extract key numerical parameters from the text content and convert the key numerical parameters into structured data.

[0054] In this embodiment of the invention, a Natural Language Processing (NLP) engine is activated. This engine first uses Optical Character Recognition (OCR) technology to scan a standard document and extract the text content from the unstructured data.

[0055] The natural language processing engine analyzes the extracted text content, automatically identifying and extracting key numerical parameters, such as capacity electricity price (yuan / kW / year) and discharge subsidy electricity price (yuan / kWh), through predefined rules and keyword matching algorithms (such as regular expressions). The engine then transforms this unstructured data into structured constraint parameters that can be directly used by mathematical models.

[0056] By combining optical character recognition technology, predefined rules, and keyword matching algorithms, unstructured data is converted into structured data, making it suitable for subsequent machine algorithms while ensuring the accuracy of the data conversion.

[0057] Step S202: Use machine learning clustering algorithms to cluster historical electricity price data in all structured data to identify typical daily scenarios.

[0058] Specifically, step S202 includes: Step S2021: The K-means clustering algorithm is used to perform unsupervised learning on the historical electricity price data in the structured data. The optimization objective is to minimize the sum of squared errors of all sample points within a cluster, and data with similar electricity price trends are aggregated into different clusters.

[0059] Step S2022: Sort the center vectors of each cluster according to the time series to generate the daily electricity price curve.

[0060] Step S2023 involves automatically matching descriptive labels to the daily electricity price curve to determine typical daily scenarios.

[0061] In this embodiment of the invention, historical electricity price data within a preset time period (e.g., the past year, 8760 hours) is extracted from a local time-series database, and the historical electricity price data is preprocessed, including data cleaning, outlier removal, and time series alignment, to ensure the continuity and accuracy of the data and lay the foundation for subsequent cluster analysis.

[0062] By calling historical electricity price data and using the K-Means machine learning clustering algorithm for unsupervised learning, the K-Means machine learning clustering algorithm optimizes historical electricity price data by minimizing the sum of squared errors (SSE) of all sample points within a cluster. It aggregates hourly data with similar electricity price trends into different clusters, thereby identifying typical electricity price fluctuation patterns.

[0063] After clustering, based on the centroid of each cluster, several typical daily electricity price curves are generated by sorting them according to time series. These curves represent different market operation modes and are given descriptive labels, such as "peak electricity price day", "valley electricity price day", and "off-peak electricity price day", to determine typical daily scenarios.

[0064] It is understandable that the elbow rule, contour coefficient or other evaluation methods can be used to analyze the clustering effect under different numbers of clusters (K value) and automatically determine the optimal number of typical daily scenes, so as to avoid the situation where the scene division is too coarse or too fine due to manually setting the K value.

[0065] By automatically identifying typical daily scenarios from historical electricity price data, it abandons the traditional simplified and fixed operational strategy assumptions and can match the optimal operating mode for pumped storage power stations based on real and dynamically changing market price signals.

[0066] Step S203: Automatically match the mixed-integer linear programming model framework corresponding to the typical daily scenario, and output the same number of parameterized mathematical optimization model instances to be solved as the typical daily scenario.

[0067] Please see details Figure 1 Step S103 of the illustrated embodiment will not be described again here.

[0068] Step S204: Use the solver to solve the parameterized mathematical optimization model instance and output the optimized scheduling curve of the pumped storage power station.

[0069] Specifically, step S204 includes: Step S2041: With minimizing the overall operating cost of the pumped storage power station as the objective function and hydraulic constraints, physical constraints, electricity market coupling constraints, and relevant document boundary constraints as constraints, solve the parameterized mathematical optimization model instance to obtain a time series array, and use the time series array to construct the optimal scheduling curve of the pumped storage power station.

[0070] In this embodiment of the invention, an objective function is defined, which is to minimize the overall operating cost of the pumped storage power station, or equivalently, to maximize the overall revenue of the pumped storage power station. The cost or revenue comprehensively considers the power generation revenue, pumping cost, standby service revenue, and start-up and shutdown costs, etc.

[0071] Embed at least the following types of constraints: Hydraulic constraints: Establish a reservoir water balance equation to ensure that the reservoir water level changes in accordance with physical laws during the scheduling cycle and are always maintained within the preset upper and lower safety limits; Physical constraints: Define the physical boundaries of the unit's operation, including but not limited to: maximum and minimum limits on power generation / pumping capacity; binary variables used to represent the unit's start-up and shutdown status; ramp constraints that limit the rate of power change; and minimum continuous start-up / shutdown time constraints to ensure stable unit operation. Market coupling constraints: Establish a mathematical relationship between power generation and the provided upward and downward reserve capacity to ensure that the sum of the three does not exceed the maximum available output of the unit, thus meeting the joint optimization requirements of the power market; Relevant document boundary constraints: Parameters such as subsidy assessment indicators are extracted from policy documents and embedded into the model as hard constraints or soft constraints (violation of soft constraints will trigger penalty costs) to ensure the compliance of scheduling strategies.

[0072] Use commercial solvers (such as CPLEX, Gurobi) or open-source solvers to solve the completed MILP model in order to find the optimal solution that satisfies all constraints.

[0073] By minimizing the overall operating cost of pumped storage power plants and applying multiple constraints, including the synergistic effect of hydraulic, physical, and market-coupled constraints, the generated optimal scheduling curve is made to conform to both physical equipment capabilities and market rules. This addresses the core pain points in both theory and practice. By solving a parameterized mathematical optimization model instance, the optimal scheduling curve of the pumped storage power plant is obtained, which can effectively improve accuracy compared to traditional calculations based on experience or simple rules.

[0074] Step S205: Using power production simulation based on unit combination and economic dispatch, compare the differences in technical indicators between the benchmark scenario and the evaluation scenario.

[0075] Specifically, step S205 includes: Step S2051: The scenario that does not include pumped storage power stations is taken as the baseline scenario, and the scenario that includes pumped storage power stations and has the same power grid model and boundaries as the baseline scenario is taken as the evaluation scenario.

[0076] Step S2052: Run the production simulation of the benchmark scenario to obtain the benchmark technical indicators.

[0077] Step S2053: Run the production simulation of the evaluation scenario to obtain the evaluation technical indicators.

[0078] Step S2054: Perform item-by-item difference calculation on the benchmark technical indicators and the evaluation technical indicators to obtain the technical indicator differences between the benchmark scenario and the evaluation scenario.

[0079] In this embodiment of the invention, a power production simulation mechanism based on Unit Commitment (UC) and Economic Dispatch (ED) is initiated, and a power grid topology model and a database of technical and economic parameters for various power sources (such as thermal power units, wind farms, photovoltaic power plants, etc.) are loaded. Load forecast data and renewable energy output forecast data for future dispatch cycles are input to complete the environmental preparation before simulation.

[0080] like Figure 2 As shown, the baseline scenario that does not include pumped storage power stations is defined as Scenario A. Pumped storage power stations are automatically removed from the power grid model or their output is reduced to zero to ensure that the simulation environment is not affected.

[0081] The production simulation for Scenario A executes unit combination and economic dispatch calculations sequentially to determine unit start-up and shutdown plans and output allocation for each time period. After the simulation, a set of benchmark technical indicators is automatically recorded and stored, including but not limited to: total power generation cost, wind and solar curtailment (MWh), total number of start-ups and shutdowns of thermal power units, and spinning reserve capacity.

[0082] Under the same grid model and boundary conditions as Scenario A, the pumped storage power station is reconnected to the grid model, and corresponding technical parameters, such as capacity and efficiency curves, are configured for it. The evaluation scenario is defined as Scenario B.

[0083] In Scenario B, the operation mode of the pumped storage power station in the future scheduling cycle will be forcibly locked to the optimized scheduling curve. This means that in subsequent production simulations, the power generation / pumping power and reserve capacity of the power station will be strictly implemented according to this curve.

[0084] In the production simulation of operation scenario B, under the constraint of the established operating curve of pumped storage power station, the unit combination and economic dispatch optimization of other power sources (such as thermal power units) are re-performed.

[0085] After the simulation, the evaluation technical indicators that are exactly the same as those in Scenario A are recorded and stored. The technical indicators of Scenario A and Scenario B are then calculated by differential calculation to obtain the quantified technical benefit increments, such as the reduction in wind and solar power curtailment, the reduction in total system power generation cost, and the reduction in the number of start-ups and shutdowns of thermal power units.

[0086] Based on differential calculation, a technology contribution assessment report is automatically generated. This report quantitatively reveals how optimized dispatch strategies improve renewable energy consumption, reduce operating costs, reduce thermal power unit wear, and improve the overall economic efficiency and stability of the power grid. The assessment results are output in the form of structured data and charts.

[0087] By running production simulations of the baseline and evaluation scenarios using the same power grid model and boundaries, the obtained evaluation technical indicators are differentially calculated to avoid interference from other factors and provide objective and reliable data for power grid dispatch.

[0088] Step S206: Collect the differences in optimized scheduling curves and technical indicators, and generate a pre-decision report.

[0089] Specifically, step S206 includes: Step S2061: Load the decision report template.

[0090] Step S2062: Display the optimized scheduling curve in the core control instruction area, and based on the differences in technical indicators, display the expected technical benefits of the scheduling strategy in the technical benefit analysis area, and generate a pre-decision report.

[0091] In this embodiment of the invention, a result aggregation program is initiated, a structured data storage area for storing and integrating data is established, the optimized scheduling curve and technical indicator differences generated in the aforementioned steps are received, the optimized scheduling curve is imported into the storage unit in the structured data storage area, and the technical indicator differences are imported into the storage unit in the structured data storage area associated with the optimized scheduling curve.

[0092] The comprehensive decision-making report generation mechanism is activated, and a preset decision-making report template is loaded. This template includes a core control instruction area, a technical benefit analysis area, and an economic benefit forecast area.

[0093] Based on the imported optimized scheduling curve data, the power scheduling curve of the pumped storage power station for the next 24 hours is generated in the core control instruction display area of ​​the decision report template in the form of charts and data tables. This curve can be directly sent to the power station control system (such as LSC, Local Control System) as an operation reference or setting reference.

[0094] Based on the differences in imported technical indicators, the technical benefit analysis area of ​​the decision report template quantitatively displays the expected technical benefits of the scheduling strategy, such as "the number of start-ups and shutdowns of thermal power units that can be reduced by the expected additional wind power absorption".

[0095] By combining technical benefit data and acquired market price data, the various benefits are estimated in monetary terms in the economic benefit forecast area of ​​the decision-making report template. For example, the benefits are calculated based on the amount of wind power consumed and the corresponding electricity price, and the benefits are calculated based on the reduction of the number of start-ups and shutdowns of thermal power units and the start-up and shutdown costs of thermal power units, so as to provide reference data for investment return analysis.

[0096] The report templates with pre-filled content for each area are integrated to generate a complete comprehensive decision-making report. This report is output in a viewable, printable, and transmittable format for relevant decision-makers to refer to.

[0097] The multi-dimensional dispatching pre-decision-making method for pumped storage power stations provided in this embodiment displays the expected technical benefits of optimized dispatching curves and dispatching strategies in the core control instruction area and technical benefit analysis area of ​​the decision report template, respectively. It adopts a visualization combined with a hierarchical design to provide data reference for decision-makers.

[0098] The multi-factor dispatch pre-decision method for pumped storage power plants has the following beneficial effects: (1) For the first time, the three dimensions of subsidy income (policy value), market participation income (transaction value), and power system operation income (system value) are unified under one framework for quantification, which can more comprehensively reflect the multiple identities and values ​​of pumped storage power stations as market entities and system regulation tools, and avoid underestimating their contributions. Through the innovative "system contribution income calculation" module, the system-level contributions of pumped storage power stations to "promoting the consumption of renewable energy" and "reducing the operating costs of thermal power units" that are difficult to price directly are successfully quantified into specific income figures. This allows their environmental and social benefits to be intuitively reflected in the final income report, providing strong data support for the comprehensive value assessment of the power station. (2) Improved the accuracy and authenticity of revenue calculation, making decision-making more reliable. It abandons the simplified and fixed operational strategy assumptions of traditional methods. Through intelligent pattern recognition, it can match the optimal operating mode for pumped storage power stations based on real and dynamically changing market price signals. By constructing a mixed integer linear programming (MILP) model, it meticulously characterizes complex physical constraints such as reservoir capacity, unit efficiency, frequency regulation, and energy coupling. Its solution results are far more accurate than calculations based on experience or simple rules, and better reflect the optimal decision-making behavior and potential revenue of the power station in a real market environment. (3) It possesses the ability to dynamically adapt to policy and market environments, and its assessment methods are more forward-looking and timely. Through NLP technology, it can automatically track and analyze the latest government subsidy policies and quickly integrate them into the calculation model, greatly improving the response speed and assessment timeliness to policy changes; (4) The calculation process has been automated and the decision support has been intelligentized, significantly improving the evaluation efficiency and practicality.

[0099] This embodiment also provides a multi-element dispatch pre-decision-making device for pumped-storage power stations. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0100] This embodiment provides a multi-element dispatch pre-decision device for pumped storage power stations, such as... Figure 3 As shown, it includes: The acquisition module 301 is used to acquire raw structured data related to electricity and unstructured data related to standard documents, and convert the unstructured data into structured data. The raw structured data and the structured data obtained from the conversion of unstructured data together constitute all the structured data. Clustering module 302 is used to cluster historical electricity price data in all structured data using machine learning clustering algorithms to identify typical daily scenarios, including peak days, average days, and low-end days. Matching module 303 is used to automatically match the mixed integer linear programming model framework corresponding to the typical daily scene and output the same number of parameterized mathematical optimization model instances to be solved as the typical daily scene. Solver module 304 is used to solve the parameterized mathematical optimization model instance using the solver and output the optimized scheduling curve of pumped storage power station.

[0101] In some optional implementations, the acquisition module 301 includes: The first extraction unit is used to extract text content from unstructured data using optical character recognition technology; The second extraction unit is used to extract key numerical parameters from the text content using predefined rules and keyword matching algorithms, and to convert the key numerical parameters into structured data. The key numerical parameters include capacity electricity price.

[0102] In some alternative implementations, clustering module 302 includes: The aggregation unit is used to perform unsupervised learning on historical electricity price data in all structured data using the K-means clustering algorithm. The optimization objective is to minimize the sum of squared errors of all sample points within a cluster, and to aggregate data with similar electricity price trends into different clusters. The sorting unit is used to sort the center vectors of each cluster according to the time series to generate the daily electricity price curve; The determination unit is used to automatically match descriptive labels to the daily electricity price curve and determine typical daily scenarios. The descriptive labels include peak electricity price days, off-peak electricity price days, and low-peak electricity price days.

[0103] In some alternative implementations, the solver module 304 includes: The solution unit is used to solve a parameterized mathematical optimization model instance with the objective function of minimizing the overall operating cost of the pumped storage power station and with hydraulic constraints, physical constraints, electricity market coupling constraints, and relevant document boundary constraints as constraints. The result is a time series array, which is used to construct the optimal scheduling curve of the pumped storage power station.

[0104] In some alternative implementations, hydraulic constraints include reservoir water balance equations, physical constraints include physical boundaries of unit operation, and market coupling constraints include ensuring that the sum of generating power, increased reserve capacity, and decreased reserve capacity does not exceed the unit's maximum available output.

[0105] In some alternative embodiments, the device further includes: Production simulation module 305 is used to compare the differences in technical indicators between the benchmark scenario and the evaluation scenario by using power production simulation based on unit combination and economic dispatch. The report generation module 306 is used to collect the differences in optimization scheduling curves and technical indicators to generate a pre-decision report.

[0106] In some alternative implementations, the production simulation module 305 includes: The scenario determination unit is used to take a scenario that does not include a pumped storage power station as the baseline scenario and a scenario that includes a pumped storage power station and has the same power grid model and boundaries as the baseline scenario as the evaluation scenario. The first production simulation unit is used to run production simulations of benchmark scenarios to obtain benchmark technical indicators. The second production simulation unit is used to run production simulations of the evaluation scenario to obtain evaluation technical indicators. The differential calculation unit is used to perform differential calculations on the benchmark technical indicators and the evaluation technical indicators one by one to obtain the technical indicator differences between the benchmark scenario and the evaluation scenario.

[0107] In some alternative implementations, the report generation module 306 includes: The loading unit is used to load the decision report template, which includes a core control instruction area and a technical benefit analysis area. The report generation unit is used to display the optimized scheduling curve in the core control instruction area, and based on the differences in technical indicators, to display the expected technical benefits of the scheduling strategy in the technical benefit analysis area, and generate a pre-decision report.

[0108] The multi-factor dispatch pre-decision device for pumped-storage power stations provided in this embodiment of the invention can execute the multi-factor dispatch pre-decision method for pumped-storage power stations provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units described above are the same as in the corresponding embodiments described above, and will not be repeated here.

[0109] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0110] The following is a detailed reference. Figure 4 This diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 401, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 402 or a program loaded from memory 408 into random access memory (RAM) 403. The RAM 403 also stores various programs and data required for the operation of the electronic device. The processor 401, ROM 402, and RAM 403 are interconnected via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0111] Typically, the following devices can be connected to I / O interface 405: input devices 406 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 407 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 408 including, for example, magnetic tapes, hard disks, etc.; and communication devices 409. Communication device 409 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 4 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0112] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 409, or installed from a memory 408, or installed from a ROM 402. When the computer program is executed by the processor 401, it performs the functions defined in the multi-dispatch pre-decision method for pumped-storage power stations according to embodiments of the present invention.

[0113] Figure 4The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0114] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the multi-factor dispatch pre-decision method for pumped-storage power stations shown in the above embodiments is implemented.

[0115] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0116] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and all such modifications and variations fall within the scope defined by the appended invention.

Claims

1. A multi-element scheduling pre-decision method for pumped storage power stations, characterized by, The method includes: Acquire raw structured data related to electricity and unstructured data related to standard documents, and convert the unstructured data into structured data. The raw structured data and the structured data obtained from the conversion of unstructured data together constitute all the structured data. The historical electricity price data in all the structured data is clustered using machine learning clustering algorithms to identify typical daily scenarios, including peak days, average days, and low days. Automatically match the mixed-integer linear programming model framework corresponding to the typical daily scenario, and output the same number of parameterized mathematical optimization model instances to be solved as the typical daily scenario; The parameterized mathematical optimization model instance is solved using a solver, and the optimized scheduling curve of the pumped storage power station is output.

2. The method of claim 1, wherein, The step of converting the unstructured data into structured data includes: The text content in the unstructured data is extracted using optical character recognition technology; Key numerical parameters are extracted from the text content using predefined rules and keyword matching algorithms, and then the key numerical parameters are converted into structured data, including capacity electricity price.

3. The method of claim 1, wherein, The process of using machine learning clustering algorithms to cluster historical electricity price data from all structured data to identify typical daily scenarios includes: The K-means clustering algorithm is used to perform unsupervised learning on the historical electricity price data in all the structured data. The optimization objective is to minimize the sum of squared errors of all sample points within a cluster, and data with similar electricity price trends are aggregated into different clusters. The center vectors of each cluster are sorted according to the time series to generate the daily electricity price curve; Descriptive labels are automatically matched to the daily electricity price curve to determine typical daily scenarios. The descriptive labels include peak electricity price days, off-peak electricity price days, and low-peak electricity price days.

4. The method of claim 1, wherein, The process of solving the parameterized mathematical optimization model instance using a solver and outputting the optimized scheduling curve of the pumped storage power station includes: With the objective function of minimizing the overall operating cost of pumped storage power stations, and with hydraulic constraints, physical constraints, electricity market coupling constraints, and relevant document boundary constraints as constraints, the parameterized mathematical optimization model instance is solved to obtain a time series array, which is then used to construct the optimized scheduling curve of the pumped storage power station.

5. The method of claim 4, wherein, The hydraulic constraints include the reservoir water balance equation, the physical constraints include the physical boundaries of unit operation, and the market coupling constraints include ensuring that the sum of power generation, increased reserve capacity, and decreased reserve capacity does not exceed the maximum available output of the unit.

6. The method of claim 1, wherein, After solving the parameterized mathematical optimization model instance using a solver and outputting the optimal dispatch curve for the pumped storage power station, the method further includes: By using power production simulation based on unit combination and economic dispatch, the differences in technical indicators between the baseline scenario and the evaluation scenario are compared. The differences between the optimized scheduling curves and the technical indicators are collected to generate a pre-decision report.

7. The method of claim 6, wherein, The method utilizes power production simulation based on unit combination and economic dispatch to compare the differences in technical indicators between the baseline scenario and the evaluation scenario, including: The scenario without pumped storage power stations is used as the baseline scenario, and the scenario with pumped storage power stations and the same power grid model and boundaries as the baseline scenario is used as the evaluation scenario. Run a production simulation of a benchmark scenario to obtain benchmark technical indicators; Production simulations of the evaluation scenarios are run to obtain evaluation technical indicators; The differences between the technical indicators of the benchmark scenario and the evaluation scenario are obtained by performing item-by-item difference calculations on the benchmark technical indicators and the evaluation technical indicators.

8. The method of claim 6, wherein, The process of collecting the optimized scheduling curves and the differences in the technical indicators to generate a pre-decision report includes: Load the decision report template, which includes a core control instruction area and a technical benefit analysis area; The optimized scheduling curve is displayed in the core control instruction area. Based on the differences in technical indicators, the expected technical benefits of the scheduling strategy are displayed in the technical benefit analysis area, and a pre-decision report is generated.

9. A multi-element dispatch pre-decision device for pumped storage power stations, characterized by, The device includes: The acquisition module is used to acquire raw structured data related to electricity and unstructured data related to standard documents, and convert the unstructured data into structured data. The raw structured data and the structured data obtained from the conversion of unstructured data together constitute all the structured data. The clustering module is used to cluster the historical electricity price data in all the structured data using machine learning clustering algorithms to identify typical daily scenarios, including peak days, average days, and low-end days. The matching module is used to automatically match the mixed integer linear programming model framework corresponding to the typical daily scene and output the same number of parameterized mathematical optimization model instances to be solved as the typical daily scene. The solver module is used to solve the parameterized mathematical optimization model instance using a solver and output the optimized scheduling curve of the pumped storage power station.

10. The apparatus of claim 9, wherein, The acquisition module includes: The first extraction unit is used to extract the text content from the unstructured data using optical character recognition technology; The second extraction unit is used to extract key numerical parameters from the text content using predefined rules and keyword matching algorithms, and to convert the key numerical parameters into structured data, including capacity electricity price.

11. The apparatus of claim 9, wherein, The clustering module includes: The aggregation unit is used to perform unsupervised learning on historical electricity price data in structured data using the K-means clustering algorithm. The optimization objective is to minimize the sum of squared errors of all sample points within a cluster, and to aggregate data with similar electricity price trends into different clusters. The sorting unit is used to sort the center vectors of each cluster according to the time series to generate the daily electricity price curve; The determining unit is used to automatically match descriptive labels to the daily electricity price curve and determine typical daily scenarios, wherein the descriptive labels include peak electricity price days, off-peak electricity price days, and low-peak electricity price days.

12. The apparatus according to claim 9, characterized in that, The solution module includes: The solution unit is used to solve the parameterized mathematical optimization model instance with the objective function of minimizing the comprehensive operating cost of the pumped storage power station and with hydraulic constraints, physical constraints, electricity market coupling constraints, and relevant document boundary constraints as constraints, to obtain a time series array, which is then used to construct the optimal scheduling curve of the pumped storage power station.

13. The apparatus of claim 12, wherein, The hydraulic constraints include the reservoir water balance equation, the physical constraints include the physical boundaries of unit operation, and the market coupling constraints include ensuring that the sum of power generation, increased reserve capacity, and decreased reserve capacity does not exceed the maximum available output of the unit.

14. The apparatus of claim 9, wherein, The device further includes: The production simulation module is used to compare the differences in technical indicators between the baseline scenario and the evaluation scenario by using power production simulation based on unit combination and economic dispatch. The report generation module is used to collect the differences in optimized scheduling curves and technical indicators, and generate a pre-decision report.

15. The apparatus of claim 14, wherein, The production simulation module includes: The scenario determination unit is used to take a scenario that does not include a pumped storage power station as the baseline scenario and a scenario that includes a pumped storage power station and has the same power grid model and boundaries as the baseline scenario as the evaluation scenario. The first production simulation unit is used to run production simulations of benchmark scenarios to obtain benchmark technical indicators. The second production simulation unit is used to run production simulations of the evaluation scenario to obtain evaluation technical indicators. The differential calculation unit is used to perform differential calculations on the benchmark technical indicators and the evaluation technical indicators item by item to obtain the technical indicator differences between the benchmark scenario and the evaluation scenario.

16. The apparatus of claim 14, wherein, The report generation module includes: A loading unit is used to load a decision report template, which includes a core control instruction area and a technical benefit analysis area. The report generation unit is used to display the optimized scheduling curve in the core control instruction area, and based on the differences in technical indicators, to display the expected technical benefits of the scheduling strategy in the technical benefit analysis area, and generate a pre-decision report.

17. An electronic device, comprising: include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the multi-factor dispatch pre-decision method for pumped storage power stations as described in any one of claims 1 to 8.

18. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the multi-factor dispatch pre-decision method for pumped storage power stations as described in any one of claims 1 to 8.

Images