Method for analyzing and evaluating self-balancing capability of local power grid

Abstract

The invention discloses a local power grid self-balancing capability analysis and evaluation method. The method comprises the following steps: S1, data preparation and uncertainty scene construction: the randomness of new energy and load is quantified into a typical scene set which can be processed by a calculation model and has probability characteristics; s2, constructing a multi-dimensional collaborative optimization model: establishing an optimization model capable of coordinating actions of sources, networks, loads and storages under different time scales, and embedding market signals; s3, comprehensive evaluation and simulation analysis: operating the model, calculating evaluation indexes, and quantifying the self-balancing capability from multiple dimensions; s4, decision support and feedback optimization: interpreting an evaluation result, forming a management strategy, and continuously improving the model; the method has the advantages of being accurate in evaluation result, high in uncertainty handling capacity and high in coordination capacity, the new energy real-time consumption rate of the local power grid and the overall operation stability of the system can be greatly improved, and economic benefits are improved.

Description

Technical Field

[0001] This invention belongs to the technical field of local power grid self-balancing capacity analysis and evaluation, and specifically relates to a method for analyzing and evaluating the self-balancing capacity of a local power grid. Background Technology

[0002] Local grid self-balancing can effectively reduce the need to purchase electricity from the upper-level grid. Especially during peak electricity consumption periods, by fully exploring and utilizing local power generation resources, the operating pressure on transmission and distribution facilities can be significantly reduced, equipment overload can be avoided, equipment lifespan can be extended, and thus the operation and maintenance costs of the power grid can be reduced.

[0003] However, with the large-scale integration of distributed power sources and the rapid growth of new loads, the distribution network as a whole faces unprecedented challenges. Due to the intermittency and volatility of distributed power sources, and the spatiotemporal uncertainty of electric vehicle charging loads, the power flow of the distribution network is becoming increasingly complex and volatile, leading to frequent problems such as local grid overload and voltage exceeding limits. These issues seriously threaten the stability of local power supply, mainly manifesting in the following three aspects:

[0004] 1. Traditional deterministic models cannot handle the strong stochasticity of new energy sources and loads, and the assessment results are seriously inaccurate under extreme weather conditions, which is insufficient to cope with uncertainty;

[0005] 2. The generation, grid, load, and storage components are difficult to coordinate on a temporal and spatial scale, resulting in low overall operational efficiency;

[0006] 3. Advanced evaluation technologies (such as probabilistic analysis and AI prediction) are difficult to drive resources and play a role in practice if they are separated from market mechanisms and management rules. Technology and management mechanisms need to be integrated.

[0007] To address the aforementioned issues, it is essential to develop a method for analyzing and evaluating the self-balancing capacity of local power grids. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method for analyzing and evaluating the self-balancing capacity of a local power grid that has accurate evaluation results, strong ability to cope with uncertainties, and strong coordination capabilities. This method can significantly improve the real-time absorption rate of new energy sources in the local power grid and the overall operational stability of the system, as well as improve economic benefits.

[0009] The objective of this invention is achieved as follows: a method for analyzing and evaluating the self-balancing capability of a local power grid, comprising the following steps:

[0010] S1, Data Preparation and Uncertainty Scenario Construction: Quantifying the randomness of new energy sources and loads into a set of typical scenarios with probabilistic characteristics that can be processed by computational models;

[0011] S2, Multi-dimensional Collaborative Optimization Model Construction: Establish an optimization model that can coordinate the actions of sources, networks, loads, and storage at different time scales, and embed market signals;

[0012] S3, Comprehensive Evaluation and Simulation Analysis: Run the model, calculate evaluation indicators, and quantify self-balancing capability from multiple dimensions;

[0013] S4, Decision Support and Feedback Optimization: Interpret the evaluation results, formulate management strategies, and continuously improve the model.

[0014] Preferably, step S1 includes the following steps:

[0015] S11, Multi-temporal and spatial data acquisition and governance: Collect at least one year of historical data within the assessment area, with a time resolution of 15 minutes or 1 hour. The collected data includes historical output data of wind power and photovoltaic power, historical data of various loads, grid structure, line capacity and transformer capacity, as well as the capacity, power and charging and discharging efficiency of existing energy storage. After cleaning, alignment and labeling, a standardized historical database is formed.

[0016] S12, generate a typical daily operating scenario:

[0017] S121. Considering dynamic time warping and alignment shape, the K-means clustering algorithm is improved. The improved K-means clustering algorithm is used to perform cluster analysis on the joint data of "wind and solar load".

[0018] S122, based on the clustering results, select several of the most representative typical daily scenarios;

[0019] S123, calculate the annual probability of occurrence for each typical daily scenario, including the number of days in that category and the total number of days;

[0020] S13, Constructing a set of extreme scenarios and uncertainties:

[0021] S131, Extreme Scenario: Identify extreme combinations in historical data and use them as stress test scenarios with a probability of 1%;

[0022] S132, Uncertainty Set: For the prediction error of each typical scenario, based on the deviation between historical predictions and measured data, its fluctuation range is constructed to form a robust optimized uncertainty set.

[0023] Preferably, in step S12, typical daily scenarios include sunny days with high load in winter, cloudy and rainy days with low load in summer, windy days with low load in spring, scorching days with high load in summer, cold wave days with high load in winter, windy days with low load in spring / autumn, sunny days with low load in spring / autumn, and low load days during the Spring Festival / National Day holidays; in step S13, extreme scenarios include extremely hot and windless days, extremely cold and dark days, extremely low load days with abundant wind and sunlight in spring, consecutive weeks of cloudy and rainy days with no wind, and days hit by typhoons / floods.

[0024] Preferably, step S2 includes the following steps:

[0025] S21, Define the multi-objective optimization function: Establish the multi-objective optimization function and use the weighted sum method to convert it into a single objective;

[0026] S22, Set model constraints: Write power balance constraints, equipment operation constraints, network and security constraints, and system stability constraints into the optimization model as mathematical equations;

[0027] S23, Design a multi-timescale coordination mechanism: Establish a hierarchical optimization framework of day-ahead scheduling, intraday rolling optimization, and real-time control;

[0028] S24, Embedding Market Mechanisms and Price Signals: Integrating economic signals into technical models.

[0029] Preferably, in step S21, the multi-objective optimization function formula is:

[0030] ;

[0031] The components are defined as follows:

[0032] (1) Total system operating cost: ;

[0033] In the formula, Cost of exchanging data with the mainnet;

[0034] In the formula, Conventional power generation costs;

[0035] In the formula, Load reduction incurs penalty costs;

[0036] In the formula, Energy storage depreciation costs;

[0037] (2) Overall power imbalance of the system: ;

[0038] (3) Total carbon emissions of the system: ;

[0039] The meanings of the symbols in the aforementioned formula are as follows:

[0040] s and S are the scene index and the total number of scenes, respectively; Let be the probability of scenario s occurring; These are the weighting coefficients for economic efficiency, stability, and environmental friendliness, respectively.

[0041] t and T are the time period index and the total number of time periods, respectively; The unit time period length (hours);

[0042] and The electricity purchase and sales prices (yuan / kWh) are respectively for time period t. and These are the power consumption for purchasing and selling electricity (kW), respectively.

[0043] g and G are the index and total number of conventional units, respectively; The active power output (kW) of unit g; Let g be the cost coefficient of unit g; This represents the start / stop status of unit g, where 1 indicates that unit g is running and 0 indicates that unit g is stopped.

[0044] d and D are the load reduction index and total number, respectively; The penalty cost for reducing load d (yuan / kWh); The power reduction (kW) for load d;

[0045] b and B are the energy storage device index and total number, respectively; The unit charge-discharge depreciation cost of energy storage b (RMB / kW); and These are the charging and discharging power (kW) of energy storage b, respectively.

[0046] The system net unbalanced power (kW); Total load demand (kW);

[0047] The carbon emission intensity of unit g is (kg-CO2 / kWh).

[0048] Preferably, in step S22, the model constraints are as follows:

[0049] (1) Power balance constraint:

[0050] In the formula, w and W are the index and total number of new energy generating units, respectively;

[0051] (2) Equipment operating constraints:

[0052] Conventional unit constraints: (upper and lower limits of output); (Climb rate constraint);

[0053] Constraints of new energy units: In the formula, Maximum predicted output of new energy (kW);

[0054] Constraints of energy storage systems:

[0055] ; ; (Charging and discharging power limits); where, and These represent the energy storage charging and discharging states, with 1 indicating charging and 0 indicating not charging or discharging. and These are the maximum charging and discharging power of the energy storage (kW), respectively.

[0056] ; (Energy state constraints); where, The energy state (kWh) of energy storage b at the end of time period t; and These refer to the energy storage charging and discharging efficiencies, respectively. and These are the lower and upper limits of energy storage (kWh), respectively.

[0057] Flexible load constraints: ; In the formula, This represents the maximum instantaneous reduction ratio of the load d; This represents the maximum daily total reduction percentage for load d.

[0058] (3) Network and security constraints:

[0059] (Line power flow constraints); where l is the line index; The transmission power limit of line l (kW);

[0060] (Node voltage constraints); where n is the node index; The voltage amplitude at node n (kV);

[0061] (4) System stability constraints:

[0062] (Primary frequency modulation standby constraint); where, and Spinning reserve capacity (kW) provided for generating units and energy storage, respectively; This is the system's backup demand coefficient.

[0063] Preferably, in step S23, the hierarchical optimization framework is as follows:

[0064] (1) Day-ahead scheduling:

[0065] Model: Run the complete MILP model of steps S21 and S22, with the input being a typical daily scene;

[0066] Output: Unit start-up and shutdown plan Mainnet exchange plan and Energy storage energy state reference curve ;

[0067] (2) Intraday rolling optimization:

[0068] Model: With fixed unit start-up and shutdown states, and using updated ultra-short-term forecasts as input, the model re-optimizes unit output within a shorter time window. and energy storage plans and ;

[0069] Output: Adjusted power plan, correcting for day-ahead deviations;

[0070] (3) Real-time control:

[0071] Model: Employing rule-based control or model predictive control to respond to real-time unbalanced power in the second-to-minute range. ;

[0072] Output: Issues real-time control commands to AGC units, energy storage, and rapid excitation systems.

[0073] Preferably, step S24 includes the following steps:

[0074] (1) Price signal driven: in cost In China, electricity retailers Time-of-use pricing and retail electricity price It can be different from this to reflect market supply and demand;

[0075] (2) Demand response modeling: Setting price trigger conditions for flexible loads, ,in The benchmark electricity price;

[0076] (3) Simulated flexibility market: In the model, energy storage and interruptible loads are treated as flexibility resources, and their call-up costs are... and This can be considered as its price quote, and the system will clear it according to a uniform pricing method to assess market efficiency.

[0077] Preferably, step S3 includes the following steps:

[0078] S31, Run optimization simulation: Simulate all scenarios built in step S1. As input, these values ​​are sequentially substituted into the multi-dimensional collaborative optimization model established in step S2 for calculation, and the simulation results for each scenario s are output as follows:

[0079] (1) Processing sequence of each device: , , , ;

[0080] (2) Exchange power sequence with the main network: , ;

[0081] (3) Load reduction sequence: ;

[0082] (4) System operating costs: ;

[0083] (5) Various auxiliary variables: , ;

[0084] S32, Calculate multi-dimensional evaluation indicators: Based on the simulation results of step S31, calculate four types of evaluation indicators, as follows:

[0085] (1) Coordination and optimization indicators:

[0086] Electrical imbalance: In the formula, The power imbalance in scenario s. This represents the system's net unbalanced power.

[0087] New energy consumption rate: In the formula, The renewable energy consumption rate under scenario s;

[0088] Multi-energy integrated coordination coefficient: In the formula, Let be the coordination coefficient under scenario s. ;

[0089] (2) Safety and stability indicators:

[0090] Probability of voltage exceeding limit: In the formula, Let N be the probability that the node voltage exceeds the safety limit in scenario s, and N be the total number of nodes.

[0091] Frequency pass rate: In the formula, For frequency pass rate, For system frequency, For the rated frequency, This is the frequency deviation limit;

[0092] (3) Economic benefit indicators:

[0093] Levelized cost per unit of electricity: In the formula, The total operating cost in scenario s. This represents the total investment cost after being converted to an annualized value.

[0094] Average daily operating cost: ;

[0095] (4) Social benefit indicators:

[0096] CO2 emission reduction per unit of electricity:

[0097] In the formula, For reference scenarios, carbon emissions This represents the carbon emissions under the current operating mode.

[0098] Finally, output an S×M evaluation metric matrix X, where S is the number of scenes, M is the evaluation metric, and the matrix elements are... This represents the value of the m-th indicator in scenario s;

[0099] S33, Determine the indicator weights and overall score:

[0100] (1) Objective weighting; using the entropy weighting method;

[0101] Indicator Standardization: For the evaluation matrix X, calculate the standardized value of the sm-th indicator: (For benefit-oriented indicators); (For cost-related indicators);

[0102] Calculate information entropy: ;

[0103] Calculate the weights: ;

[0104] (2) Subjective weights; The Delphi method was used to summarize the subjective weight vectors of each indicator through multiple rounds of expert questionnaires. ;

[0105] (3) Combined weight values: The subjective and objective weights are linearly combined to obtain the final weight; In the formula, This refers to the objective weighting preference coefficient.

[0106] (4) Overall score: a weighted average is used to determine the legality. ;

[0107] Final output: Comprehensive score of local power grid self-balancing capability: And the scores for each dimension: the indicators are divided into subsets according to dimensions, and the scores for coordination and optimization, safety and stability, economic benefits and social benefits are calculated separately.

[0108] Preferably, step S4 includes the following steps:

[0109] S41, Shortcomings Analysis and Risk Assessment: Analyze dimensions with low overall scores, identify system weaknesses, generate a "System Vulnerability Analysis Report," and clearly point out weak links and potential risks;

[0110] S42, propose optimization suggestions and investment plans: In response to the shortcomings, propose technical solutions and market mechanism design suggestions, and generate the "Local Power Grid Self-Balancing Capacity Improvement Plan".

[0111] S43, Recommended feedback loop: Compare actual operating data with evaluation and prediction data, periodically re-execute steps S1, S2 and S3, update typical scenarios and optimize model parameters, and achieve self-improvement of the evaluation system.

[0112] Due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0113] (1) This invention combines probabilistic scenario generation with multi-stage robust optimization to construct typical and extreme scenarios, thereby quantifying uncertainty into calculable risk probabilities. This transforms the evaluation conclusion from a single, deterministic point estimate into a result distribution that includes risk boundaries, greatly improving the resilience and planning reliability of the system in the face of fluctuations and extreme events. It overcomes the problem that traditional deterministic models cannot handle the strong randomness of new energy sources and loads, and the evaluation results are seriously inaccurate under extreme weather conditions.

[0114] (2) This invention establishes a multi-timescale coordination and scheduling framework of "day-to-day-real-time" and constructs a unified optimization model of source-grid-load-storage to achieve global coordination at the technical level, thereby breaking the barrier of independent optimization of each element. Based on hierarchical control and dynamic rolling optimization, it achieves power balance across time scales, greatly improves the real-time absorption rate of new energy and the overall operational stability of the system, and solves the problem that the existing source, grid, load and storage links are difficult to coordinate in time and space scales, resulting in low overall operating efficiency.

[0115] (3) This invention embeds the market mechanism as the core variable into the technology optimization model and conducts a comprehensive evaluation through a combination of subjective and objective weighting methods. This enables the pure technology model to simulate the response behavior of resource subjects under economic signals, ensuring that the optimization strategy is not only technically feasible but also economically driving. This greatly enhances the practical feasibility and decision-making value of the evaluation results and solves the problem that existing advanced evaluation technologies are difficult to drive resources and play a role in practice if they are separated from market mechanisms and management rules.

[0116] In summary, this invention has the advantages of accurate evaluation results, strong ability to cope with uncertainties, and strong coordination capabilities. It can significantly improve the real-time absorption rate of new energy in local power grids and the overall operational stability of the system, and improve economic benefits. Attached Figure Description

[0117] Figure 1 This is a flowchart of the method steps of the present invention. Detailed Implementation

[0118] The technical solution of the present invention will be further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0119] like Figure 1 As shown, the present invention provides a method for analyzing and evaluating the self-balancing capability of a local power grid, comprising the following steps:

[0120] S1, Data Preparation and Uncertainty Scenario Construction: Quantify the randomness of new energy sources and loads into a set of typical scenarios with probabilistic characteristics that can be processed by computational models.

[0121] S11, Multi-temporal Data Acquisition and Governance: Collect at least one year of historical data within the assessment area, with a time resolution of 15 minutes or 1 hour. The collected data includes:

[0122] Source side: Historical power output data for wind and solar power;

[0123] Load side: Historical data for various load types, such as industrial, commercial, and residential;

[0124] Power grid: grid structure, line capacity, transformer capacity;

[0125] Storage side: Existing energy storage capacity, power, and charge / discharge efficiency;

[0126] Step S11 output: A cleaned, aligned and labeled standardized historical database.

[0127] S12, generate a typical daily operating scenario:

[0128] S121, an improved K-means clustering algorithm (considering dynamic time warping and alignment shape) is used to perform clustering analysis on the joint "wind and solar load" data;

[0129] S122. Based on the clustering results, select 5-8 of the most representative typical day scenarios, such as sunny days with high load in winter, cloudy and rainy days with slender lotus in summer, windy days with low load in spring, hot days with high load in summer, cold wave days with high load in winter, windy days with slender lotus in spring / autumn, sunny days with slender lotus in spring / autumn, and low load days during the Spring Festival / National Day holiday.

[0130] S123, calculate the annual probability of occurrence for each typical daily scenario (number of days in this category / total number of days).

[0131] Step S12 output: a set of typical daily scene curves and their corresponding occurrence probabilities.

[0132] S13, Constructing a set of extreme scenarios and uncertainties:

[0133] S131, Extreme Scenarios: Identify extreme combinations in historical data, such as extremely hot and windless, extremely cold and dark, extremely low load in spring with days of abundant wind and sunshine, consecutive weeks of cloudy and windless weather, and days hit by typhoons / floods, and use them as stress test scenarios with a probability of 1%.

[0134] S132, Uncertainty Set: For the prediction error of each typical scenario, based on the deviation between historical predictions and measured data, its fluctuation range (such as ±3σ confidence interval) is constructed to form a robust optimized uncertainty set.

[0135] Step S13 outputs: extreme scenarios for risk verification and uncertainty ranges for robust optimization.

[0136] S2, Multi-dimensional Collaborative Optimization Model Construction: Establish an optimization model that can coordinate the actions of sources, networks, loads, and storage at different time scales, and embed market signals.

[0137] S21, Define the multi-objective optimization function: Establish the multi-objective optimization function and use the weighted sum method to convert it into a single objective. The formula for the multi-objective optimization function is:

[0138] ;

[0139] The components are defined as follows:

[0140] (1) Total system operating cost: ;

[0141] In the formula, Cost of exchanging data with the mainnet;

[0142] In the formula, Conventional power generation costs;

[0143] In the formula, Load reduction incurs penalty costs;

[0144] In the formula, Energy storage depreciation costs;

[0145] (2) Overall power imbalance of the system: ;

[0146] (3) Total carbon emissions of the system: ;

[0147] The meanings of the symbols in the aforementioned formula are as follows:

[0148] s and S are the scene index and the total number of scenes, respectively; Let be the probability of scenario s occurring; The weighting coefficients are respectively for economic efficiency, stability, and environmental friendliness, and ;

[0149] t and T are the time period index and the total number of time periods, respectively; The unit time period length (hours);

[0150] and The electricity purchase and sales prices (yuan / kWh) are respectively for time period t. and These are the power consumption for purchasing and selling electricity (kW), respectively.

[0151] g and G are the index and total number of conventional units, respectively; The active power output (kW) of unit g; Let g be the cost coefficient of unit g; This represents the start / stop status of unit g, where 1 indicates that unit g is running and 0 indicates that unit g is stopped.

[0152] d and D are the load reduction index and total number, respectively; The penalty cost for reducing load d (yuan / kWh); The power reduction (kW) for load d;

[0153] b and B are the energy storage device index and total number, respectively; The unit charge-discharge depreciation cost of energy storage b (RMB / kW); and These are the charging and discharging power (kW) of energy storage b, respectively.

[0154] The system net unbalanced power (kW); Total load demand (kW);

[0155] The carbon emission intensity of unit g is (kg-CO2 / kWh).

[0156] S22, Set model constraints: Write power balance constraints, equipment operation constraints, network and security constraints, and system stability constraints into the optimization model using mathematical equations. The model constraints are as follows:

[0157] (1) Power balance constraint:

[0158] In the formula, w and W are the index and total number of new energy generating units, respectively;

[0159] (2) Equipment operating constraints:

[0160] Conventional unit constraints: (upper and lower limits of output); (Climb rate constraint);

[0161] Constraints of new energy units: In the formula, Maximum predicted output of new energy (kW);

[0162] Constraints of energy storage systems:

[0163] ; ; (Charging and discharging power limits); where, and These represent the energy storage charging and discharging states, with 1 indicating charging and 0 indicating not charging or discharging. and These are the maximum charging and discharging power of the energy storage (kW), respectively.

[0164] ; (Energy state constraints); where, The energy state (kWh) of energy storage b at the end of time period t; and These refer to the energy storage charging and discharging efficiencies, respectively. and These are the lower and upper limits of energy storage (kWh), respectively.

[0165] Flexible load constraints: ; In the formula, This represents the maximum instantaneous reduction ratio of the load d; This represents the maximum daily total reduction percentage for load d.

[0166] (3) Network and security constraints:

[0167] (Line power flow constraints); where l is the line index; The transmission power limit of line l (kW);

[0168] (Node voltage constraints); where n is the node index; The voltage amplitude at node n (kV);

[0169] (4) System stability constraints:

[0170] (Primary frequency modulation standby constraint); where, and Spinning reserve capacity (kW) provided for generating units and energy storage, respectively; This is the system's backup demand coefficient.

[0171] S23, Design a multi-timescale coordination mechanism: Establish a hierarchical optimization framework of day-ahead scheduling, intraday rolling optimization, and real-time control. The hierarchical optimization framework is as follows:

[0172] (1) Day-ahead scheduling:

[0173] Model: Run the complete MILP model of steps S21 and S22, with the input being a typical daily scene;

[0174] Output: Unit start-up and shutdown plan Mainnet exchange plan and Energy storage energy state reference curve ;

[0175] (2) Intraday rolling optimization:

[0176] Model: With fixed unit start-up and shutdown states, and using updated ultra-short-term forecasts as input, the model re-optimizes unit output within a shorter time window. and energy storage plans and ;

[0177] Output: Adjusted power plan, correcting for day-ahead deviations;

[0178] (3) Real-time control:

[0179] Model: Employing rule-based control or model predictive control to respond to real-time unbalanced power in the second-to-minute range. ;

[0180] Output: Issues real-time control commands to AGC units, energy storage, and rapid excitation systems.

[0181] S24, Embedding Market Mechanisms and Price Signals: Integrating economic signals into technical models, including the following steps:

[0182] (1) Price signal driven: in cost In China, electricity retailers Time-of-use pricing and retail electricity price It can be different from this to reflect market supply and demand;

[0183] (2) Demand response modeling: Setting price trigger conditions for flexible loads, ,in The benchmark electricity price;

[0184] (3) Simulated flexibility market: In the model, energy storage and interruptible loads are treated as flexibility resources, and their call-up costs are... and This can be considered as its price quote, and the system will clear it according to a uniform pricing method to assess market efficiency.

[0185] S3, Comprehensive Evaluation and Simulation Analysis: Run the model, calculate evaluation indicators, and quantify self-balancing capability from multiple dimensions.

[0186] S31, Run optimization simulation: Simulate all scenarios built in step S1. As input, including typical daily scenarios and extreme scenarios, these are sequentially substituted into the multi-dimensional collaborative optimization model established in step S2 for solution calculation, and the simulation results for each scenario s are output as follows:

[0187] (1) Processing sequence of each device: , , , ;

[0188] (2) Exchange power sequence with the main network: , ;

[0189] (3) Load reduction sequence: ;

[0190] (4) System operating costs: ;

[0191] (5) Various auxiliary variables: , .

[0192] S32, Calculate multi-dimensional evaluation indicators: Based on the simulation results of step S31, calculate four types of evaluation indicators, as follows:

[0193] (1) Coordination and optimization indicators:

[0194] Electrical imbalance: In the formula, The power imbalance in scenario s. This represents the system's net unbalanced power.

[0195] New energy consumption rate: In the formula, The renewable energy consumption rate under scenario s;

[0196] Multi-energy integrated coordination coefficient: In the formula, Let be the coordination coefficient under scenario s. .

[0197] (2) Safety and stability indicators:

[0198] Probability of voltage exceeding limit: In the formula, Let N be the probability that the node voltage exceeds the safety limit in scenario s, and N be the total number of nodes.

[0199] Frequency pass rate: In the formula, For frequency pass rate, For system frequency, For the rated frequency, This is the frequency deviation limit.

[0200] (3) Economic benefit indicators:

[0201] Levelized cost per unit of electricity: In the formula, The total operating cost in scenario s. This represents the total investment cost after being converted to an annualized value.

[0202] Average daily operating cost: .

[0203] (4) Social benefit indicators:

[0204] CO2 emission reduction per unit of electricity:

[0205] In the formula, For reference scenarios, carbon emissions This represents the carbon emissions under the current operating mode.

[0206] Finally, output an S×M evaluation metric matrix X, where S is the number of scenes, M is the evaluation metric, and the matrix elements are... This represents the value of the m-th indicator in scenario s.

[0207] S33, Determine the indicator weights and overall score:

[0208] (1) Objective weighting; using the entropy weighting method;

[0209] Indicator Standardization: For the evaluation matrix X, calculate the standardized value of the sm-th indicator: (For benefit-oriented indicators); (For cost-related indicators);

[0210] Calculate information entropy: ;

[0211] Calculate the weights: .

[0212] (2) Subjective weights; The Delphi method was used to summarize the subjective weight vectors of each indicator through multiple rounds of expert questionnaires. .

[0213] (3) Combined weight values: The subjective and objective weights are linearly combined to obtain the final weight; In the formula, This represents the objective weighting preference coefficient.

[0214] (4) Overall score: a weighted average is used to determine the legality. .

[0215] Final output: Comprehensive score of local power grid self-balancing capability: And the scores for each dimension: the indicators are divided into subsets according to dimensions, and the scores for coordination and optimization, safety and stability, economic benefits and social benefits are calculated separately.

[0216] S4, Decision Support and Feedback Optimization: Interpret the evaluation results, formulate management strategies, and continuously improve the model.

[0217] S41, Shortcomings Analysis and Risk Assessment: Analyze dimensions with low overall scores, identify system weaknesses, generate a "System Vulnerability Analysis Report," and clearly point out weak links and potential risks;

[0218] S42, propose optimization suggestions and investment plans: In response to the shortcomings, propose technical solutions and market mechanism design suggestions, and generate the "Local Power Grid Self-Balancing Capacity Improvement Plan".

[0219] S43, Recommended feedback loop: Compare actual operating data with evaluation and prediction data, periodically re-execute steps S1, S2 and S3, update typical scenarios and optimize model parameters, and achieve self-improvement of the evaluation system.

[0220] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of the claims of the present invention.

Claims

1. A method for analyzing and evaluating the self-balancing capacity of a local power grid, characterized in that, Includes the following steps: S1, Data Preparation and Uncertainty Scenario Construction: Quantifying the randomness of new energy sources and loads into a set of typical scenarios with probabilistic characteristics that can be processed by computational models; S2, Multi-dimensional Collaborative Optimization Model Construction: Establish an optimization model that can coordinate the actions of sources, networks, loads, and storage at different time scales, and embed market signals; S3, Comprehensive Evaluation and Simulation Analysis: Run the model, calculate evaluation indicators, and quantify self-balancing capability from multiple dimensions; S4, Decision Support and Feedback Optimization: Interpret the evaluation results, formulate management strategies, and continuously improve the model.

2. The method for analyzing and evaluating the self-balancing capacity of a local power grid according to claim 1, characterized in that, Step S1 includes the following steps: S11, Multi-temporal and spatial data acquisition and governance: Collect at least one year of historical data within the assessment area, with a time resolution of 15 minutes or 1 hour. The collected data includes historical output data of wind power and photovoltaic power, historical data of various loads, grid structure, line capacity and transformer capacity, as well as the capacity, power and charging and discharging efficiency of existing energy storage. After cleaning, alignment and labeling, a standardized historical database is formed. S12, generate a typical daily operating scenario: S121. Considering dynamic time warping and alignment shape, the K-means clustering algorithm is improved. The improved K-means clustering algorithm is used to perform cluster analysis on the joint data of "wind and solar load". S122, based on the clustering results, select several of the most representative typical daily scenarios; S123, calculate the annual probability of occurrence for each typical daily scenario, including the number of days in that category and the total number of days; S13, Constructing a set of extreme scenarios and uncertainties: S131, Extreme Scenario: Identify extreme combinations in historical data and use them as stress test scenarios with a probability of 1%; S132, Uncertainty Set: For the prediction error of each typical scenario, based on the deviation between historical predictions and measured data, its fluctuation range is constructed to form a robust optimized uncertainty set.

3. The method for analyzing and evaluating the self-balancing capability of a local power grid according to claim 2, characterized in that: In step S12, typical daily scenarios include sunny days with high load in winter, cloudy and rainy days with low load in summer, windy days with low load in spring, scorching days with high load in summer, cold wave days with high load in winter, windy days with low load in spring / autumn, sunny days with low load in spring / autumn, and low load days during the Spring Festival / National Day holidays. In step S13, extreme scenarios include extremely hot and windless days, extremely cold and dark days, extremely low load days with abundant wind and sunlight in spring, consecutive weeks of cloudy and rainy days with no wind, and days hit by typhoons / floods.

4. The method for analyzing and evaluating the self-balancing capacity of a local power grid according to claim 1, characterized in that, Step S2 includes the following steps: S21, Define the multi-objective optimization function: Establish the multi-objective optimization function and use the weighted sum method to convert it into a single objective; S22, Set model constraints: Write power balance constraints, equipment operation constraints, network and security constraints, and system stability constraints into the optimization model as mathematical equations; S23, Design a multi-timescale coordination mechanism: Establish a hierarchical optimization framework of day-ahead scheduling, intraday rolling optimization, and real-time control; S24, Embedding Market Mechanisms and Price Signals: Integrating economic signals into technical models.

5. The method for analyzing and evaluating the self-balancing capability of a local power grid according to claim 4, characterized in that, In step S21, the multi-objective optimization function formula is: ； The components are defined as follows: (1) Total system operating cost: ; In the formula, Cost of exchanging data with the mainnet; In the formula, The cost of conventional power generation; In the formula, Load reduction incurs penalty costs; In the formula, Energy storage depreciation costs; (2) Overall power imbalance of the system: ; (3) Total carbon emissions of the system: ; The meanings of the symbols in the aforementioned formula are as follows: s and S are the scene index and the total number of scenes, respectively; Let be the probability of scenario s occurring; These are the weighting coefficients for economic efficiency, stability, and environmental friendliness, respectively. t and T are the time period index and the total number of time periods, respectively; The unit time period length (hours); and The electricity purchase and sales prices (yuan / kWh) are respectively for time period t. and These are the power purchased and the power sold (kW), respectively. g and G are the index and total number of conventional units, respectively; The active power output (kW) of unit g; Let g be the cost coefficient of unit g; This represents the start / stop status of unit g, where 1 indicates that unit g is running and 0 indicates that unit g is stopped. d and D are the load reduction index and total number, respectively; The penalty cost for reducing load d (yuan / kWh); The power reduction (kW) for load d; b and B are the energy storage device index and total number, respectively; The unit charge-discharge depreciation cost of energy storage b (RMB / kW); and These are the charging and discharging power (kW) of energy storage b, respectively. The system net unbalanced power (kW); Total load demand (kW); The carbon emission intensity of unit g is (kg-CO2 / kWh).

6. The method for analyzing and evaluating the self-balancing capacity of a local power grid according to claim 4, characterized in that, In step S22, the model constraints are as follows: (1) Power balance constraint: In the formula, w and W are the index and total number of new energy generating units, respectively; (2) Equipment operating constraints: Conventional unit constraints: (upper and lower limits of output); (Climb rate constraint); Constraints of new energy units: In the formula, Maximum predicted output of new energy (kW); Constraints of energy storage systems: ; ; (Charging and discharging power limits); where, and These represent the energy storage charging and discharging states, with 1 indicating charging and 0 indicating not charging or discharging. and These are the maximum charging and discharging power of the energy storage (kW), respectively. ; (Energy state constraints); where, The energy state (kWh) of energy storage b at the end of time period t; and These refer to the energy storage charging and discharging efficiencies, respectively. and These are the lower and upper limits of energy storage (kWh), respectively. Flexible load constraints: ; In the formula, This represents the maximum instantaneous reduction ratio of the load d; This represents the maximum daily total reduction percentage for load d. (3) Network and security constraints: (Line power flow constraints); where l is the line index; The transmission power limit of line l (kW); (Node voltage constraints); where n is the node index; The voltage amplitude at node n (kV); (4) System stability constraints: (Primary frequency modulation standby constraint); where, and Spinning reserve capacity (kW) provided for generating units and energy storage, respectively; This is the system's backup demand coefficient.

7. The method for analyzing and evaluating the self-balancing capacity of a local power grid according to claim 4, characterized in that, In step S23, the hierarchical optimization framework is as follows: (1) Day-ahead scheduling: Model: Run the complete MILP model of steps S21 and S22, with the input being a typical daily scene; Output: Unit start-up and shutdown plan Mainnet exchange plan and Energy storage energy state reference curve ; (2) Intraday rolling optimization: Model: With fixed unit start-up and shutdown states, and using updated ultra-short-term forecasts as input, the model re-optimizes unit output within a shorter time window. and energy storage plans and ; Output: Adjusted power plan, correcting for day-ahead deviations; (3) Real-time control: Model: Employing rule-based control or model predictive control to respond to real-time unbalanced power in the second-to-minute range. ; Output: Issues real-time control commands to AGC units, energy storage, and rapid excitation systems.

8. The method for analyzing and evaluating the self-balancing capacity of a local power grid according to claim 4, characterized in that, Step S24 includes the following steps: (1) Price signal driven: in cost In China, electricity retailers Time-of-use pricing and retail electricity price It can be different from this to reflect market supply and demand; (2) Demand response modeling: Setting price trigger conditions for flexible loads, ,in The benchmark electricity price; (3) Simulated flexibility market: In the model, energy storage and interruptible loads are treated as flexibility resources, and their call-up costs are... and This can be considered as its price quote, and the system will clear it according to a uniform pricing method to assess market efficiency.

9. The method for analyzing and evaluating the self-balancing capacity of a local power grid according to claim 1, characterized in that, Step S3 includes the following steps: S31, Run optimization simulation: Simulate all scenarios built in step S1. As input, these values ​​are sequentially substituted into the multi-dimensional collaborative optimization model established in step S2 for calculation, and the simulation results for each scenario s are output as follows: (1) Processing sequence of each device: , , , ; (2) Exchange power sequence with the main network: , ; (3) Load reduction sequence: ; (4) System operating costs: ; (5) Various auxiliary variables: , ; S32, Calculate multi-dimensional evaluation indicators: Based on the simulation results of step S31, calculate four types of evaluation indicators, as follows: (1) Coordination and optimization indicators: Electrical imbalance: In the formula, The power imbalance in scenario s. This represents the system's net unbalanced power. New energy consumption rate: In the formula, The renewable energy consumption rate under scenario s; Multi-energy integrated coordination coefficient: In the formula, Let be the coordination coefficient under scenario s. ; (2) Safety and stability indicators: Probability of voltage exceeding limit: In the formula, Let N be the probability that the node voltage exceeds the safety limit in scenario s, and N be the total number of nodes. Frequency pass rate: In the formula, For frequency pass rate, For system frequency, For the rated frequency, This is the frequency deviation limit; (3) Economic benefit indicators: Levelized cost per unit of electricity: In the formula, The total operating cost in scenario s. This represents the total investment cost after being converted to an annualized value. Average daily operating cost: ; (4) Social benefit indicators: Carbon dioxide emission reduction per unit of electricity: In the formula, For reference scenarios, carbon emissions This represents the carbon emissions under the current operating mode. Finally, output an S×M evaluation metric matrix X, where S is the number of scenes, M is the evaluation metric, and the matrix elements are... This represents the value of the m-th indicator in scenario s; S33, Determine the indicator weights and overall score: (1) Objective weighting; using the entropy weighting method; Indicator Standardization: For the evaluation matrix X, calculate the standardized value of the sm-th indicator: (For benefit-oriented indicators); (For cost-related indicators); Calculate information entropy: ; Calculate the weights: ; (2) Subjective weighting; Using the Delphi method, subjective weight vectors of each indicator were obtained through multiple rounds of expert questionnaires. ; (3) Combined weight values: The subjective and objective weights are linearly combined to obtain the final weight; In the formula, This refers to the objective weighting preference coefficient. (4) Overall score: a weighted average is used to determine the legality. ; Final output: Comprehensive score of local power grid self-balancing capability: And the scores for each dimension: the indicators are divided into subsets according to dimensions, and the scores for coordination and optimization, safety and stability, economic benefits and social benefits are calculated separately.

10. The method for analyzing and evaluating the self-balancing capacity of a local power grid according to claim 1, characterized in that, Step S4 includes the following steps: S41, Shortcomings Analysis and Risk Assessment: Analyze dimensions with low overall scores, identify system weaknesses, generate a "System Vulnerability Analysis Report," and clearly point out weak links and potential risks; S42, propose optimization suggestions and investment plans: In response to the shortcomings, propose technical solutions and market mechanism design suggestions, and generate the "Local Power Grid Self-Balancing Capacity Improvement Plan". S43, Recommended feedback loop: Compare actual operating data with evaluation and prediction data, periodically re-execute steps S1, S2 and S3, update typical scenarios and optimize model parameters, and achieve self-improvement of the evaluation system.

Images