Multi-time and space coupling double-layer optimization configuration method for integrated wind-solar-hydro-storage base

Abstract

This invention relates to the field of power system technology, specifically disclosing a multi-temporal coupled two-layer optimization configuration method for integrated wind, solar, hydro, and energy storage bases. The method includes data preprocessing and spatiotemporal feature extraction; collecting wind, solar, and hydropower generation data, load data, grid data, and energy storage system data, and performing data cleaning and outlier processing on the collected data; then standardizing the processed data; constructing a grid risk assessment system; constructing a two-layer model for energy storage optimization configuration and operation; constructing an upper-layer energy storage planning model; considering the grid risk assessment system, constructing a lower-layer energy storage operation model; and solving the two-layer model for energy storage optimization configuration and operation based on mixed integer programming to obtain the configuration scheme. The advantages of this invention are that it can achieve global optimization of energy storage configuration and operation, improving the regional grid supply guarantee level and clean energy consumption rate while reducing system operating costs and enhancing the creativity and practicality of the solution.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a multi-temporal and spatial coupling dual-layer optimization configuration method for integrated wind, solar, hydro and storage bases. Background Technology

[0002] High-proportion clean energy power grids with weak transmission ends face challenges such as insufficient external transmission channels and limited flexible resources. On the one hand, large-scale clean energy generation can easily lead to wind and solar curtailment, hindering the achievement of "dual carbon" goals. On the other hand, high-penetration clean energy reduces the operating capacity of traditional synchronous generator units, weakening the grid's regulation capabilities and increasing the risk of power shortages, creating a dual contradiction of "curtailment" and "power shortage." Existing energy storage optimization research often focuses on a single supply guarantee or consumption target, making it difficult to balance both. Furthermore, it lacks targeted risk characterization and quantitative analysis of the actual needs of regional power grids. Additionally, it suffers from problems such as crude data processing, insufficient model adaptability to real-world scenarios, difficulty in balancing solution efficiency and accuracy, and a lack of dynamic correction mechanisms, resulting in poor optimization outcomes. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art and provide a multi-temporal and spatial coupling dual-layer optimization configuration method for integrated wind, solar, water and storage bases.

[0004] The objective of this invention is achieved through the following technical solution: a multi-temporal and spatial coupling dual-layer optimization configuration method for integrated wind, solar, hydro, and storage bases, the method comprising,

[0005] Data preprocessing and spatiotemporal feature extraction; collecting data on wind, solar, and hydropower generation, load data, power grid data, and energy storage system data, and performing data cleaning and outlier handling on the collected data.

[0006] Then, the processed data is standardized.

[0007] Construct a power grid risk assessment system, which includes power grid balance risk modeling, clean energy consumption risk modeling, load supply guarantee risk modeling, and risk level classification and dynamic early warning threshold determination.

[0008] Construct a two-layer model for energy storage optimization configuration and operation; it includes constructing an upper-layer energy storage planning model with the minimum overall cost of the entire system as the objective function; and constructing a lower-layer energy storage operation model with the minimum operating cost of the system as the objective function.

[0009] The configuration scheme is obtained by solving a two-level model of energy storage optimization configuration and operation based on mixed integer programming.

[0010] Specifically, the data cleaning and outlier handling includes identifying outliers in the data using box plots and filling in occasional outliers using linear interpolation; for continuous outlier data segments, correction or replacement is performed by combining equipment operation logs and environmental data; and for missing data, short-term missing data is filled in using the average of data at adjacent time points, while long-term missing data is filled in using a time series prediction model based on an LSTM neural network.

[0011] Specifically, the data is standardized using the Z-score standardization method:

[0012] ;

[0013] In the formula, The original data, The mean of the data. This represents the standard deviation of the data.

[0014] Specifically, the power grid balance risk modeling is as follows:

[0015] ;

[0016] In the formula, It is a synchronous generator set; , , These are the maximum, minimum, and actual output power of synchronous generator unit a, respectively. This represents the system's positive reserve capacity. This represents the system's negative reserve capacity. and These represent positive and negative backup risks, respectively. This represents the total maximum load of the system.

[0017] The risk model for clean energy consumption is as follows:

[0018] ;

[0019] In the formula, Risks associated with the consumption of clean energy; The number of energy storage systems; This refers to the number of synchronous generator sets; , These refer to the number of clean energy generating units and the number of load nodes, respectively. The power generation capacity of clean energy unit d; The maximum load power of load node c; Let b be the actual output power of the energy storage system; when At that time, the system faces the risk of power curtailment; the higher the value, the greater the risk of power curtailment.

[0020] Load supply risk modeling is as follows:

[0021] ;

[0022] In the formula, , These are risk weights for power shortage and power shortage duration, respectively, to meet the requirements. ; The actual output power of the clean energy unit d; H represents the duration of power outage; H represents the dispatch cycle.

[0023] Specifically, the risk levels are divided into four levels: low risk, medium risk, high risk, and extremely high risk. Based on extreme scenarios and system safety operation boundaries in historical data, 1,000 sets of scenario data are generated using the Monte Carlo simulation method. The risk value distribution under different scenarios is calculated, and combined with the upper limit of economic losses that the system can withstand, the warning thresholds of each risk indicator are dynamically determined.

[0024] Specifically, the objective function of the upper-level energy storage planning model is:

[0025] ;

[0026] In the formula, This represents the total cost of the energy storage system. The system's power generation / purchase cost;

[0027] ;

[0028] In the formula, This is the energy storage conversion factor; The unit capacity investment cost for energy storage is expressed in yuan / MWh. The rated energy storage capacity for node l is MWh; The unit operating cost of energy storage is expressed in yuan / MW. Let be the actual operating power of the energy storage at node l, in MW; and r be the discount rate. For energy storage lifespan; The number of energy storage systems;

[0029] ;

[0030] ;

[0031] ;

[0032] ;

[0033] In the formula, For water and electricity costs; For photovoltaic costs; For the cost of the connecting line; This refers to the unit cost of hydropower. , The quantities are respectively hydropower and photovoltaic power. , Let K be the power output of hydropower (k) and J be the power output of photovoltaic power (j), respectively, in MW. , These represent the unit switching cost of the tie line and the unit cost of photovoltaic power, respectively, in yuan / MW; The net switching power of the tie line is expressed in MW.

[0034] Specifically, the constraints of the upper-level energy storage planning model include:

[0035] Constraints on the generation side:

[0036] ;

[0037] In the formula, , These are the maximum outputs of photovoltaic power j and hydropower power k, respectively, in MW;

[0038] Grid-side constraints:

[0039] ;

[0040] In the formula, The line transmission power at node ab is in MW. , Let a and b be the phase angles, respectively, in rad. Let be the reactance of line ab, in Ω; The maximum transmission power of line ab is MW; The maximum transmission power of the tie line is MW;

[0041] Energy storage side constraints:

[0042] ;

[0043] In the formula, The rated energy storage capacity of node l is MW; The energy storage capacity of node l is MWh; 0-1 variables; The maximum energy storage capacity of node l is MW; The maximum capacity of node l is MWh; The actual operating power of the energy storage at node l is in MW;

[0044] System power balance constraints:

[0045] ;

[0046] In the formula, Let m be the power of the load node. This refers to the total number of adjustable power supplies; The output power of the i-th adjustable power source

[0047] Specifically, the objective function of the lower-level energy storage operation model is:

[0048] ;

[0049] ;

[0050] ;

[0051] ;

[0052] ;

[0053] ;

[0054] ;

[0055] In the formula, For hydropower time-series costs; Cost of electricity generated by photovoltaic power generation units; The cost of penalties for abandoning light; For the operating costs of energy storage systems; Switching power costs for tie lines; The cost of power outages; It is the penalty cost per unit of abandoned light power; and These represent the power generation and actual output of the photovoltaic generator unit at node j, respectively. Let be the power generation of the hydroelectric generator unit at node k at time t; Let be the operating power of the energy storage system at node l at time t; Let be the net switching power of the tie line at time t; Let be the load power shortage at time t.

[0056] Specifically, the constraints of the lower-level energy storage operation model include:

[0057] System power balance constraints:

[0058] ;

[0059] In the formula, Let be the output power of the i-th adjustable power source at time t. Let be the power of the m-th load node at time t, in MW.

[0060] Rotational spare constraint:

[0061] ;

[0062] In the formula, The system's spinning reserve during time period t; Let be the maximum output power of the i-th adjustable power source at time t. Let be the maximum output power of the k-th hydropower unit at time t. Let be the maximum discharge power of the l-th energy storage system at time t, in MW.

[0063] Balancing supply assurance with consumption promotion constraints:

[0064] ;

[0065] In the formula, λ1 is the minimum curtailment rate of the system due to the inability to absorb clean energy;

[0066] Power generation output constraints:

[0067] ;

[0068] Energy storage side constraints:

[0069] ;

[0070] In the formula, Let be the charging and discharging power of the energy storage system at node l at time t; and These represent the initial and final capacities of the energy storage system located at node l within a day; and These represent the minimum and maximum operating capacities of the energy storage system at node l during the scheduling process.

[0071] The present invention has the following advantages:

[0072] This invention constructs a scientific risk assessment system and a two-layer optimization model incorporating multi-temporal coupling features through key steps such as data preprocessing and spatiotemporal feature extraction, multi-dimensional verification and dynamic correction. This achieves global optimization of energy storage configuration and operation, improves the regional power grid supply level and clean energy consumption rate, reduces system operating costs, enhances the creativity and practicality of the solution, and provides more reliable technical support for the safe, stable, economical and efficient operation of high-proportion clean energy systems. Attached Figure Description

[0073] Figure 1 This is a schematic diagram of the optimized configuration method of the present invention;

[0074] Figure 2 This is a two-layer model diagram of the present invention; Detailed Implementation

[0075] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention; that is, the described embodiments are merely some embodiments of the invention, and not all embodiments. The components of the embodiments of the invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0076] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0077] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0078] The present invention will be further described below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the following description.

[0079] like Figures 1 to 2 As shown, the multi-temporal and spatial coupling dual-layer optimization configuration method for integrated wind-solar-hydro-storage bases is characterized by: the method comprising,

[0080] Data preprocessing and spatiotemporal feature extraction; collection of wind, solar, and hydropower generation data, load data, grid data, and energy storage system data. Generation data includes raw environmental data such as wind speed, solar intensity, and hydrological flow, as well as equipment parameters such as rated power, actual output, and operating efficiency of generator sets; load data includes historical load curves, load types, and maximum and minimum load values ​​for each load node; grid data includes line parameters, transmission capacity, node phase angle constraints, and tie-line switching power limits; energy storage system data includes unit capacity investment cost, operating cost, charging and discharging efficiency, rated power and capacity limits, and service life. The data collection time span is no less than 3 years, with a time resolution of 15 minutes per data collection, ensuring data integrity and timeliness.

[0081] The collected data underwent data cleaning and outlier handling. A combination of statistical analysis and machine learning was used for data processing. First, box plots were used to identify outliers, such as sudden increases or decreases in power generation exceeding equipment physical limits, or excessive deviations in load data from historical data. For occasional outliers, linear interpolation was used to fill in the gaps. For continuous outlier segments, corrections or replacements were made based on equipment operation logs and environmental data. To address data gaps, short-term gaps were filled using the average of adjacent time points, while long-term gaps were filled using a time series prediction model based on an LSTM neural network, ensuring data quality.

[0082] Spatiotemporal feature mining includes,

[0083] Time dimension: Analyze the intraday fluctuation characteristics of wind and solar power generation, such as the peak and valley changes of photovoltaic output during the day, the random fluctuation cycle of wind power output, seasonal variation patterns, such as higher wind power output in winter and sufficient photovoltaic output in summer, the peak and valley period distribution of load, such as the daytime peak of industrial load and the evening peak of residential load, and extract key time characteristic indicators, such as fluctuation coefficient, peak-valley difference rate, duration, etc.

[0084] Spatial dimension: Analyze the complementarity of wind and solar resources in different regions, such as when wind power output in region A is low, wind power output in region B is high. Analyze the spatial distribution matching degree of load nodes and power generation nodes, the distribution of transmission bottlenecks in power grid lines, construct a spatial correlation matrix, and quantify the resource synergy potential and power grid constraint strength between regions.

[0085] Then, the processed data is standardized. In order to eliminate the difference in dimensions between different types of data, the various types of data collected and processed are standardized, such as converting physical quantities such as power generation and load power into normalized values ​​relative to the rated values.

[0086] The standardization process uses the Z-score standardization method:

[0087] ;

[0088] In the formula, The original data, The mean of the data. The standard deviation of the data is set to ensure that the processed data follows a normal distribution with a mean of 0 and a variance of 1, thereby improving the convergence speed and computational accuracy of subsequent model solutions. Data preprocessing and spatiotemporal feature extraction provide high-quality data support for subsequent risk assessment and model construction.

[0089] A power grid risk assessment system is constructed. This system is used to quantify the supply and absorption risks of high-proportion clean energy systems. It includes power grid balance risk modeling, clean energy absorption risk modeling, load supply risk modeling, and risk level classification and dynamic early warning threshold determination. This invention addresses the contradiction of "power curtailment" and "power shortage" coexisting in high-proportion clean energy systems by establishing a three-level assessment system covering power grid balance risk, clean energy absorption risk, and load supply risk.

[0090] In the main power grid, grid balance risk primarily considers the system's reserve capacity. The risk indicator characterizes the grid balance risk as focusing on the problem of insufficient system reserve capacity, categorized into two types: positive reserve insufficiency and negative reserve insufficiency. A quantitative model is constructed based on the parameters of synchronous generator units and energy storage systems. The grid balance risk model is as follows:

[0091] ;

[0092] In the formula, This refers to the number of synchronous generator sets; , , These are the maximum, minimum, and actual output power of synchronous generator unit a, respectively. This represents the system's positive reserve capacity. This represents the system's negative reserve capacity. and These represent positive and negative backup risks, respectively. The system time is set to 1 resolution; This represents the total maximum load of the system.

[0093] The risk of clean energy consumption is centered on the curtailment of wind and solar power. It is characterized by the difference between power generation and load power. The model of clean energy consumption risk is as follows:

[0094] ;

[0095] In the formula, Risks associated with the consumption of clean energy; The number of energy storage systems; , , These are the number of clean energy generating units, the number of load nodes, and the number of synchronous generator units, respectively. , These are the actual output power of synchronous generator unit a and energy storage system b, respectively; The power generation capacity of clean energy unit d; The maximum load power of load node c;

[0096] Load supply risk is mainly characterized by power rationing and outages on the load side. Its severity can be comprehensively assessed by the amount of power shortage and the duration of the shortage. The risk index characterizes load supply risk by comprehensively characterizing the amount of power shortage and the duration of the shortage. The load supply risk model is as follows:

[0097] ;

[0098] In the formula, , These are risk weights for power shortage and power shortage duration, respectively, to meet the requirements. ; The actual output power of the clean energy unit d; H represents the duration of power outage; H is the scheduling cycle, which is 24 hours.

[0099] The weights of three primary risk indicators were determined using the analytic hierarchy process (AHP). Risk levels and warning thresholds were dynamically set based on Monte Carlo simulation and the upper limit of economic losses, providing a precise risk basis for subsequent optimization. The weights for grid balance risk (0.35), clean energy consumption risk (0.35), and load supply guarantee risk (0.3) were calculated. A weighted summation was used to obtain the comprehensive system risk value, classifying the risk levels into four categories: low risk (0-0.2), medium risk (0.2-0.5), high risk (0.5-0.8), and extremely high risk (0.8-1.0). Based on extreme scenarios and system safety operation boundaries from historical data, 1000 sets of scenario data were generated using Monte Carlo simulation. The risk value distribution under different scenarios was calculated, and the warning thresholds for each risk indicator were dynamically determined based on the upper limit of economic losses the system could withstand. Dynamic early warning threshold determination: Based on extreme scenarios in historical data, such as extreme weather leading to strong wind and solar power generation or sudden load increases, and the boundary of safe system operation, 1000 sets of scenario data are generated using the Monte Carlo simulation method. The risk value distribution under different scenarios is calculated. Combined with the upper limit of economic losses that the system can bear, such as the total power shortage loss and curtailment loss not exceeding 5% of the system's annual operating cost, the early warning thresholds for each risk indicator are dynamically determined. For example, the early warning threshold for grid balance risk is set to 0.4, the early warning threshold for clean energy consumption risk is set to 0.35, and the early warning threshold for load supply guarantee risk is set to 0.3. When the risk value exceeds the early warning threshold, the subsequent model's specific constraint adjustment is triggered. The grid risk assessment system is used to accurately characterize the supply guarantee and consumption risk status of a high-proportion clean energy system and provide a basis for risk early warning.

[0100] A two-layer model for energy storage optimization configuration and operation is constructed, and collaborative optimization is achieved through parameter iteration. This includes constructing an upper-layer energy storage planning model with the minimum overall cost of the entire system as the objective function, and constructing a lower-layer energy storage operation model with the minimum operating cost of the system as the objective function. The upper and lower-layer models are solved collaboratively through parameter iteration, while incorporating multi-temporal coupling characteristics and risk warning constraints in the model construction process.

[0101] The upper-level energy storage planning model aims to minimize the overall system cost by determining the energy storage configuration nodes, rated power, and capacity. This must satisfy constraints such as power generation output, grid transmission, and energy storage capacity / power, while also considering the impact of multi-temporal and spatial characteristics on the configuration. The objective function of the upper-level energy storage planning model is:

[0102] ;

[0103] In the formula, This represents the total cost of the energy storage system. The system's power generation / purchase cost;

[0104] ;

[0105] In the formula, This is the energy storage conversion factor; The unit capacity investment cost for energy storage is expressed in yuan / MWh. The rated energy storage capacity for node l is MWh; The unit operating cost of energy storage is expressed in yuan / MW. Let be the actual operating power of the energy storage at node l, in MW; and r be the discount rate. The energy storage lifespan is set at 10 years. The number of energy storage systems;

[0106] ;

[0107] ;

[0108] ;

[0109] ;

[0110] In the formula, For water and electricity costs; For photovoltaic costs; For the cost of the connecting line; This refers to the unit cost of hydropower. , The quantities are respectively hydropower and photovoltaic power. , Let K be the power output of hydropower (k) and J be the power output of photovoltaic power (j), respectively, in MW. , These represent the unit switching cost of the tie line and the unit cost of photovoltaic power, respectively, in yuan / MW; The net switching power of the tie line is expressed in MW.

[0111] Constraints on the generation side:

[0112] ;

[0113] In the formula, , These are the maximum outputs of photovoltaic power j and hydropower power k, respectively, in MW;

[0114] Grid-side constraints:

[0115] ;

[0116] In the formula, The line transmission power at node ab is in MW. , Let a and b be the phase angles, respectively, in rad. Let be the reactance of line ab, in Ω; The maximum transmission power of line ab is MW; The maximum transmission power of the tie line is MW;

[0117] Energy storage side constraints:

[0118] ;

[0119] In the formula, The rated energy storage capacity of node l is MW; The energy storage capacity of node l is MWh; 0-1 variables; The maximum energy storage capacity of node l is MW; The maximum capacity of node l is MWh; The actual operating power of the energy storage at node l is in MW;

[0120] System power balance constraints:

[0121] ;

[0122] In the formula, Let m be the power of the load node. This refers to the total number of adjustable power supplies; Let be the output power of the i-th adjustable power supply;

[0123] The lower-level optimization operation considers minimizing the system's operating cost, taking into account grid balance risk, clean energy consumption risk, and load supply guarantee risk, and transforms these risks into economic constraints. The objective function of the lower-level energy storage operation model is:

[0124] ;

[0125] ;

[0126] ;

[0127] ;

[0128] ;

[0129] ;

[0130] ;

[0131] In the formula, For hydropower time-series costs; Cost of electricity generated by photovoltaic power generation units; The cost of penalties for abandoning light; For the operating costs of energy storage systems; Switching power costs for tie lines; The cost of power outages; It is the penalty cost per unit of abandoned light power; and These represent the power generation and actual output of the photovoltaic generator unit at node j, respectively. Let be the power generation of the hydroelectric generator unit at node k at time t; Let be the operating power of the energy storage system at node l at time t; Let be the net switching power of the tie line at time t; Let be the load power shortage at time t.

[0132] The constraints of the lower-level energy storage operation model include:

[0133] System power balance constraints:

[0134] ;

[0135] In the formula, Let be the output power of the i-th adjustable power source at time t. Let be the power of the m-th load node at time t, in MW;

[0136] Rotational spare constraint:

[0137] ;

[0138] In the formula, The system's spinning reserve during time period t; Let be the maximum output power of the i-th adjustable power source at time t. Let be the maximum output power of the k-th hydropower unit at time t. Let be the maximum discharge power of the l-th energy storage system at time t, in MW;

[0139] Balancing supply assurance with consumption promotion constraints:

[0140] ;

[0141] In the formula, λ1 is the minimum curtailment rate of the system due to the inability to absorb clean energy;

[0142] Power generation output constraints:

[0143] ;

[0144] Energy storage side constraints:

[0145] ;

[0146] In the formula, Let be the charging and discharging power of the energy storage system at node l at time t; and These represent the initial and final capacities of the energy storage system located at node l within a day; and These represent the minimum and maximum operating capacities of the energy storage system at node l during the scheduling process.

[0147] The lower-level model optimizes the intraday output of wind, solar, hydro, storage, and interconnection lines, incorporating spinning reserve and specific constraints to ensure supply and consumption. The upper and lower levels work together through parameter iteration—the upper-level configuration parameters serve as the lower-level boundary conditions, and the cost and risk indicators fed back from the lower-level operation provide feedback to the upper-level adjustments, achieving global optimization of energy storage configuration and operation.

[0148] The configuration scheme is obtained by solving a two-level model of energy storage optimization configuration and operation based on mixed integer programming.

[0149] Solution algorithm selection and parameter optimization:

[0150] An improved branch-and-bound method is selected as the solution algorithm for mixed integer programming. Parameter optimization is performed to address the low efficiency of traditional algorithms.

[0151] Set an initial feasible solution generation strategy: Based on the spatiotemporal feature mining results and historical optimization data, use heuristic algorithms, such as genetic algorithms, to quickly generate better initial feasible solutions and shorten the algorithm search time.

[0152] Adjust the branching strategy: adopt a strong branching strategy to prioritize branches on variables that have a greater impact on the objective function, thereby reducing the number of invalid branches; set a pruning threshold, and prune the branch directly when the objective function value of a certain branch exceeds 1.1 times the current optimal solution, thereby improving the solution efficiency.

[0153] Optimize the constraint processing order: Prioritize the processing of core constraints such as system power balance constraints and risk warning constraints to reduce the probability of constraint conflicts.

[0154] Model solving and convergence assessment:

[0155] The two-level model is transformed into a mixed-integer linear programming equation. Preprocessed, standardized data is input, and the solution is obtained using the Gurobi solver. A convergence criterion is set: when the change in the objective function value over five consecutive iterations is less than 10^-4, and the constraint satisfaction rate reaches 99.5% or higher, the model is considered converged, and the optimal solution (energy storage configuration nodes, rated power, rated capacity, and daily output curves of each device) is output. If convergence is not achieved after more than 1000 iterations, the algorithm parameters are adjusted (e.g., increasing the pruning threshold or adjusting the initial feasible solution generation strategy), and the solution is solved again.

[0156] Multi-dimensional verification and dynamic correction;

[0157] Through multi-scenario verification and dynamic parameter adjustment, the practicality and robustness of the solution are improved, which is specifically divided into three stages:

[0158] 1. Technical performance verification:

[0159] Four basic scenarios were set up: normal scenario, wind and solar power surge scenario, load surge scenario, and equipment failure scenario, to verify the technical performance of the optimization solution:

[0160] Typical scenario: Using typical daily data, verify whether the system's power shortage rate and power curtailment rate meet the design requirements (e.g., power shortage rate less than 0.5%, power curtailment rate less than 3%).

[0161] Extreme scenarios: Simulate scenarios where wind and solar power output reaches 120% of the rated value, impact scenarios where the load suddenly increases by 30%, and failure scenarios where a single generator unit fails and shuts down, to verify the coordinated response capability of the energy storage system and other equipment, and ensure that the system power shortage rate does not exceed 1% and the power curtailment rate does not exceed 5% under extreme scenarios.

[0162] 2. Verification of the effectiveness of risk prevention and control:

[0163] By comparing the changes in the three types of risk indicators of the power grid before and after optimization, the synergistic effect of the risk assessment system and the optimization model is verified.

[0164] The calculation should determine the reduction percentage of grid balance risk, clean energy consumption risk, and load supply guarantee risk after optimization, requiring a reduction of no less than 20% for each type of risk indicator.

[0165] Simulate high-risk scenarios, such as extreme weather causing drastic fluctuations in wind and solar power output, to verify whether the model can control system risks below the warning threshold by dynamically adjusting energy storage output and configuration parameters.

[0166] Dynamic correction:

[0167] Based on multi-dimensional verification results, a dynamic parameter correction mechanism is established:

[0168] If the power shortage rate exceeds the standard in a certain scenario, adjust the reserve capacity coefficient of the energy storage configuration and the charging and discharging strategy parameters of the lower-level model;

[0169] If the convergence speed is too slow, optimize the algorithm parameters, such as the initial feasible solution generation strategy and pruning threshold.

[0170] If the power grid structure, load characteristics, or wind and solar resource conditions change during long-term operation, such as the addition of new generating units or a load increase of more than 10%, data preprocessing and spatiotemporal feature extraction should be carried out again, and model parameters and constraints should be updated to ensure the adaptability of the solution.

[0171] At the model solution level, this invention clearly defines the need to satisfy multiple constraints, including those on the generation side, grid side, energy storage side, and system power balance. A mixed-integer programming method is employed to ensure the optimality and feasibility of the solution. In terms of effect verification and economic analysis, multiple sets of comparative examples (such as scenarios without energy storage and single-objective optimization) are set up to evaluate the method from multiple dimensions, including power shortage / curtailment power, risk indicators, operating costs, and marginal costs. This not only verifies the method's technical advantages in improving supply security and absorption rates while reducing costs, but also demonstrates its economic rationality through quantitative analysis, providing technical support for the safe and economical operation of high-proportion clean energy systems.

[0172] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the present invention. Any person skilled in the art can make many possible variations and modifications to the technical solution of the present invention, or modify it into equivalent embodiments, without departing from the scope of the present invention. Therefore, any modifications, equivalent changes, and alterations made to the above embodiments based on the technology of the present invention without departing from the scope of the present invention are within the protection scope of the present invention.

Claims

1. A multi-temporal and spatial coupling dual-layer optimization configuration method for integrated wind-solar-hydro-storage bases, characterized by: The method includes, Data preprocessing and spatiotemporal feature extraction; collecting data on wind, solar and hydropower generation, load data, power grid data and energy storage system data, and performing data cleaning and outlier handling on the collected data; then standardizing the processed data; Establish a power grid risk assessment system; It includes grid balance risk, clean energy consumption risk, load supply risk, as well as risk level classification and dynamic early warning threshold determination; A two-layer model for energy storage optimization configuration and operation is constructed. This includes constructing an upper-layer energy storage planning model with the minimum overall cost of the entire system as the objective function; and constructing a lower-layer energy storage operation model with the minimum operating cost of the system as the objective function, taking into account the power grid risk assessment system. The configuration scheme is obtained by solving a two-level model of energy storage optimization configuration and operation based on mixed integer programming.

2. The multi-temporal-spatial coupling dual-layer optimization configuration method for integrated wind-solar-hydro-storage bases according to claim 1, characterized in that: The data cleaning and outlier handling process includes identifying outliers in the data using box plots, filling in occasional outliers using linear interpolation, correcting or replacing continuous outlier data segments by combining equipment operation logs and environmental data, and filling in short-term missing data by using the average of adjacent time points and long-term missing data by using a time series prediction model based on an LSTM neural network.

3. The multi-temporal-spatial coupling dual-layer optimization configuration method for integrated wind-solar-hydro-storage bases according to claim 1, characterized in that: The data was standardized using the Z-score standardization method. ； In the formula, The original data, The mean of the data. This represents the standard deviation of the data.

4. The multi-temporal-spatial coupling dual-layer optimization configuration method for integrated wind-solar-hydro-storage bases according to claim 1, characterized in that: The power grid balance risk model is as follows: ； In the formula, Number of synchronous generator sets; , , These are the maximum, minimum, and actual output power of synchronous generator unit a, respectively. This represents the system's positive reserve capacity. This represents the system's negative reserve capacity. and These represent positive and negative backup risks, respectively. This represents the total maximum load of the system. The system time is set to 1 resolution; The risk model for clean energy consumption is as follows: ； In the formula, Risks associated with the consumption of clean energy; The number of energy storage systems; , , These are the number of clean energy generating units, the number of load nodes, and the number of synchronous generator units, respectively. , These are the actual output power of synchronous generator unit a and energy storage system b, respectively; The power generation capacity of clean energy unit d; The maximum load power of load node c; Load supply risk modeling is as follows: ； In the formula, , These are risk weights for power shortage and power shortage duration, respectively, to meet the requirements. ; The actual output power of the clean energy unit d; H represents the duration of power outage; H represents the dispatch cycle.

5. The multi-temporal-spatial coupling dual-layer optimization configuration method for integrated wind-solar-hydro-storage bases according to claim 4, characterized in that: The risk levels are divided into four levels: low risk, medium risk, high risk, and extremely high risk. Based on extreme scenarios and system safety operation boundaries in historical data, 1,000 sets of scenario data are generated using the Monte Carlo simulation method. The risk value distribution under different scenarios is calculated, and the warning threshold of each risk indicator is dynamically determined in combination with the upper limit of economic loss that the system can withstand.

6. The multi-temporal-spatial coupling dual-layer optimization configuration method for integrated wind-solar-hydro-storage bases according to claim 4, characterized in that: The objective function of the upper-level energy storage planning model is: ； In the formula, This represents the total cost of the energy storage system. The system's power generation / purchase cost; ； In the formula, This is the energy storage conversion factor; The unit capacity investment cost for energy storage is expressed in yuan / MWh. The rated energy storage capacity for node l is MWh; The unit operating cost of energy storage is expressed in yuan / MW. Let be the actual operating power of the energy storage at node l, in MW; and r be the discount rate. For energy storage lifespan; The number of energy storage systems; ； ； ； ； In the formula, For water and electricity costs; This refers to the unit cost of hydropower. For photovoltaic costs; For the cost of the connecting line; , The quantities are respectively hydropower and photovoltaic power. , Let K be the power output of hydropower (k) and J be the power output of photovoltaic power (j), respectively, in MW. , These represent the unit switching cost of the tie line and the unit cost of photovoltaic power, respectively, in yuan / MW; The net switching power of the tie line is expressed in MW.

7. The multi-temporal-spatial coupling dual-layer optimization configuration method for integrated wind-solar-hydro-storage bases according to claim 6, characterized in that: The constraints of the upper-level energy storage planning model include: Constraints on the generation side: ； In the formula, , These are the maximum outputs of photovoltaic power j and hydropower power k, respectively, in MW; Grid-side constraints: ； In the formula, The line transmission power at node ab is in MW. , Let a and b be the phase angles, respectively, in rad. Let be the reactance of line ab, in Ω; The maximum transmission power of line ab is MW; The maximum transmission power of the tie line is MW; Energy storage side constraints: ； In the formula, The rated energy storage capacity of node l is MW; The energy storage capacity of node l is MWh; Variables are 0-1; The maximum energy storage capacity of node l is MW; The maximum capacity of node l is MWh; The actual operating power of the energy storage at node l is in MW; System power balance constraints: ； In the formula, Let m be the power of the load node. The total number of adjustable power supplies. Let be the output power of the i-th adjustable power source.

8. The multi-temporal-spatial coupling dual-layer optimization configuration method for integrated wind-solar-hydro-storage bases according to claim 7, characterized in that: The objective function of the lower-level energy storage operation model is: ； ； ； ； ； ； ； In the formula, For hydropower time-series costs; Cost of electricity generated by photovoltaic power generation units; The cost of penalties for abandoning light; For the operating costs of energy storage systems; Switching power costs for tie lines; The cost of power outages; It is the penalty cost per unit of abandoned light power; and These represent the power generation and actual output of the photovoltaic generator unit at node j, respectively. Let be the power generation of the hydroelectric generator unit at node k at time t; Let be the operating power of the energy storage system at node l at time t; Let be the net switching power of the tie line at time t; Let be the load power shortage at time t.

9. The multi-temporal-spatial coupling dual-layer optimization configuration method for integrated wind-solar-hydro-storage bases according to claim 8, characterized in that: The constraints of the lower-level energy storage operation model include: System power balance constraints: ； In the formula, Let be the output power of the i-th adjustable power source at time t. Let be the power of the m-th load node at time t, in MW; Rotational spare constraint: ； In the formula, The system's spinning reserve during time period t; Let be the maximum output power of the i-th adjustable power source at time t. Let be the maximum output power of the k-th hydropower unit at time t. Let be the maximum discharge power of the l-th energy storage system at time t, in MW; Balancing supply assurance with consumption promotion constraints: ； In the formula, λ1 is the minimum curtailment rate of the system due to the inability to absorb clean energy; Power generation output constraints: ； Energy storage side constraints: ； In the formula, Let be the charging and discharging power of the energy storage system at node l at time t; and These represent the initial and final capacities of the energy storage system located at node l within a day; and These represent the minimum and maximum operating capacities of the energy storage system at node l during the scheduling process.

Images