Power grid planning scene construction method and system based on power system subject behavior simulation, medium and processor

Abstract

The invention discloses a power grid planning scene construction method and system based on power system subject behavior simulation, a medium and a processor, and relates to the technical field of power system planning. The method comprises the steps of firstly simulating and analyzing power system subject behaviors to obtain subject output and load constraints; extracting a basic operation mode from the historical operation data, and forming a basic scene set through a clustering algorithm; then, the LSTM sequence prediction model is combined with environment correction to generate future operation prediction data; then, in combination with the basic scene and future data, screening key scenes covering different renewable energy permeability, load levels and extreme weather, and constructing a typical scene library; and finally, importing the typical scene into a power grid planning optimization model, and determining a final planning scheme by combining a subject behavior simulation result for simulation evaluation. The method improves the adaptability of the scene and the actual working condition, guarantees that the planning scheme gives consideration to the cost, safety and new energy consumption, and is suitable for the precise planning of a power grid.

Description

Technical Field

[0001] This invention relates to the field of power system planning technology, and in particular to a method, system, medium, and processor for constructing power grid planning scenarios based on the simulation of the main behaviors of power system entities. Background Technology

[0002] Existing methods for constructing power grid planning scenarios have significant limitations and are difficult to adapt to the development needs of power system market-oriented reforms and the high proportion of new energy grid integration.

[0003] Traditional planning scenarios often rely on experience-based settings or simple historical data statistics, failing to fully consider the autonomous behavior characteristics of multiple market players—such as the output optimization strategies of thermal power companies, the random power generation characteristics of new energy companies, and the demand response behavior of users. This leads to a disconnect between the scenario and actual operating conditions, resulting in planning schemes lacking practicality.

[0004] Meanwhile, existing methods lack sufficient accuracy in predicting future operating conditions, fail to effectively integrate time-series forecasts with environmental factor corrections, struggle to cover combined operating conditions with high, medium, and low renewable energy penetration rates and load levels, and also neglect the impact of extreme weather on system operation.

[0005] In addition, traditional scenario construction suffers from high redundancy and insufficient coverage. It fails to extract key scenarios through scientific clustering and screening mechanisms, resulting in low planning and simulation efficiency. Furthermore, the final solution often fails to balance investment costs, operational efficiency, and renewable energy consumption needs, thus failing to provide comprehensive support for the safe and stable operation of the power grid.

[0006] Therefore, there is a need for a method, system, medium, and processor for constructing power grid planning scenarios based on the simulation of the main behaviors of the power system. Summary of the Invention

[0007] To address the problems of low accuracy, disconnect from actual operating conditions, and lack of practicality in existing power grid planning schemes, this invention provides a method, system, medium, and processor for constructing power grid planning scenarios based on the simulation of the main behaviors of the power system. This improves the adaptability of the constructed scenarios to actual operating conditions, covers multi-dimensional combinations of operating conditions, and enhances the efficiency of planning simulation. The specific technical solution is as follows: A method for constructing power grid planning scenarios based on simulation of power system agent behavior includes: S1: Simulate and analyze the behavior of the main components of the power system to obtain the output and load constraints of each component; S2: Extract basic operating modes from historical power system operating data, reduce scenario redundancy through clustering algorithms, form a basic scenario set covering normal operating states, and provide underlying support for subsequent scenario expansion; S3: Using the LSTM sequence prediction model, with historical power system operation data as input, it predicts the system operation status for different time periods in the future and generates future operation prediction data; S4: Combining basic operation scenarios and future operation forecast data, select key scenarios covering high, medium and low renewable energy penetration rates, high, medium and low load levels, and the impact of extreme weather, and construct a typical scenario library for power grid planning; S5: Using scenarios from the typical scenario library as input, import them into the power grid planning optimization model, combine the simulation results of the main behavior of the power system to conduct simulation evaluation, and determine the final power grid planning scheme.

[0008] Furthermore, in step S2, the extraction of basic operating modes from historical power system operating data, and the reduction of scenario redundancy through clustering algorithms to form a basic scenario set covering normal operating states, includes the following steps: S21: Select the core feature vectors for clustering and perform preprocessing; S22: The K-means algorithm is used to cluster the core clustering feature variables; S23: Calculate the mean of the feature vectors for each cluster, convert them back to actual physical quantities, obtain the key parameters of each basic scenario, and form a set of basic scenarios. , Based on scenario K.

[0009] Furthermore, in step S3, the LSTM sequence prediction model, using historical power system operating data as input, predicts the system operating status for different future time periods to generate future operating prediction data, including the following steps: S31: Determine the input feature sequence and output prediction vector of the prediction model; S32: Construct an LSTM sequence prediction model and train it; S33: Correct the prediction results of the LSTM sequence prediction model to obtain future prediction data; the formula for correcting the prediction results is as follows: ; in, This is the revised forecast data for future operation; This is the environmental impact coefficient; for The deviation rate between real-time predicted environmental data and historical average environmental data; This is the output prediction vector of the LSTM sequence prediction model.

[0010] Furthermore, in step S4, the process of combining basic operational scenarios with future operational forecast data to screen out key scenarios covering high, medium, and low renewable energy penetration rates, high, medium, and low load levels, and the impact of extreme weather, and constructing a typical scenario library for power grid planning, includes the following steps: S41: Generate a combined scenario based on the basic scenario and future operational prediction data, using the following formula: ; In the formula, A collection of basic scenarios; These are revised future forecasts; For combined scenes; S42: Set screening criteria for screening combined scenarios, including renewable energy penetration coverage, load level coverage, extreme operating condition coverage, and scenario differences; S43: After filtering based on the screening criteria, key scenarios are retained to form a typical scenario library as follows: , Scenes that have been selected and retained.

[0011] Furthermore, in step S5, the process of using scenarios from the typical scenario library as input, importing them into the power grid planning optimization model, and combining the simulation results of the main behaviors of the power system for simulation evaluation to determine the final power grid planning scheme includes the following steps: S51: Construct a power grid planning optimization model; S52: Import each scenario into the power grid planning optimization model to obtain the corresponding optimal planning scheme, and determine the evaluation index value of the optimal planning scheme for each scenario; S53: Obtain the corresponding weighted score based on the evaluation index value of the optimal planning scheme in each scenario, and determine the final power grid planning scheme based on the weighted score.

[0012] Furthermore, the objective function of the power grid planning optimization model is as follows: ; ; ; ; in, For investment costs; Cost per unit cross-section of the circuit; This refers to the line length; The cross-sectional area of ​​the line; Cost per unit capacity of substation For substation capacity; Operating costs; The cost of fuel per unit of electricity; To maintain costs; The constraints of the power grid planning optimization model are as follows: Power balance constraints: ; Equipment capacity constraints , ; Constraints on the absorption of new energy sources: ; Subject behavior constraints: Output and load constraints of each subject; In the above formula, This is the upper limit for line transmission; This is the upper limit of substation capacity; For the first Power generation entities Always put in the effort; For the first individual users Constant load; For the first Line Time loss; , , These are the power generation entities, users, and the number of power lines, respectively. Let t be the actual amount of renewable energy power generated by the system at time t; Let t be the total power generation of the new energy power generation enterprise.

[0013] Furthermore, in step S53, the formula for determining the weighted score is as follows: ; in, For weighted scores; , These represent the minimum and maximum annual average costs for each option; , , These are the weights of the corresponding indicators; This represents the average annual cost. For the renewable energy consumption rate; The safe pass rate is N-1.

[0014] A power grid planning scenario construction system based on power system agent behavior simulation, applied to the aforementioned power grid planning scenario construction method based on power system agent behavior simulation, includes: The analysis module is used to simulate and analyze the behavior of the main components of the power system, and obtain the output and load constraints of each component. The clustering module is used to extract basic operating patterns from historical operating data of the power system, reduce scenario redundancy through clustering algorithms, form a basic scenario set covering the normal operating state, and provide underlying support for subsequent scenario expansion. The prediction module is used to use an LSTM sequence prediction model, taking historical power system operation data as input, to predict the system operation status for different time periods in the future and generate future operation prediction data. The scenario module is used to combine basic operation scenarios with future operation forecast data to select key scenarios covering high, medium and low renewable energy penetration rates, high, medium and low load levels, and the impact of extreme weather, and to build a typical scenario library for power grid planning. The simulation module is used to take scenarios from the typical scenario library as input, import them into the power grid planning and optimization model, combine the simulation results of the main behavior of the power system to conduct simulation evaluation, and determine the final power grid planning scheme.

[0015] A computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to execute the above-described method for constructing a power grid planning scenario based on simulation of the main behavior of a power system.

[0016] A processor for running a program, wherein the program executes the above-described method for constructing a power grid planning scenario based on simulation of the main behavior of a power system.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Improve the adaptability of scene construction to actual working conditions and solve the problem of incomplete characterization of main behaviors. Existing technologies often rely on experience or simple data statistics to set planning scenarios, failing to consider the differences in behavior among multiple stakeholders, such as the output optimization of thermal power plants, the stochastic generation of renewable energy plants, and user demand response. This leads to a disconnect between the scenario and reality. This application simulates and analyzes the behavior of the main stakeholders in the power system through step S1, clarifying the output and load constraints of each stakeholder (such as the upper and lower limits of thermal power plant unit output, the stochastic constraints of renewable energy output, and the rigid range of user electricity demand). It deeply integrates the behavioral characteristics of stakeholders into the entire scenario construction process, enabling the planned scenario to accurately reflect the decision-making logic and operational rules of multiple stakeholders in the actual system. This significantly improves the fit between the scenario and actual operating conditions, laying the foundation for the practicality of subsequent planning schemes.

[0018] 2. Improve the accuracy of future operating condition predictions, covering multi-dimensional combinations of operating conditions. Traditional methods lack scientific time-series models to predict future operating conditions and do not incorporate environmental factor corrections, making it difficult to cover different renewable energy penetration rates, load levels, and extreme weather conditions. Step S3 of this application employs an LSTM sequence prediction model, using historical operating data as input, and incorporates environmental impact coefficients (such as the impact of wind speed and temperature on renewable energy output) to correct the prediction results, accurately predicting the system's operating status at different future time periods. Step S4 further combines basic scenarios with corrected future prediction data to select key scenarios covering high, medium, and low renewable energy penetration rates, high / medium / low load levels, and the impact of extreme weather (typhoons, cold waves), filling the gap in multi-dimensional operating condition coverage in existing technologies and ensuring that the planning scheme can cope with complex and ever-changing future operating environments.

[0019] 3. Optimize scene quality and improve planning simulation efficiency. Existing technologies suffer from high redundancy in scenario construction and insufficient extraction of key scenarios, leading to large computational loads and low efficiency in planning simulation. This application addresses this by using the K-means clustering algorithm in step S2 to cluster basic operating patterns extracted from historical operating data, reducing scenario redundancy and forming a basic scenario set covering normal operating states. Step S4 further filters key scenarios based on clear selection criteria (renewable energy penetration coverage, load level coverage, extreme operating condition coverage, and scenario differences), constructing a concise yet comprehensive library of typical scenarios. This significantly reduces the interference of invalid scenarios on simulation calculations, improves the computational efficiency of the power grid planning optimization model, and ensures the representativeness and coverage of scenarios, avoiding the waste of planning resources due to scenario redundancy.

[0020] 4. Achieve multi-objective optimization of the planning scheme, balancing cost and operational performance. Traditional planning schemes often struggle to balance multiple objectives, including investment costs, operational efficiency, and renewable energy integration. Step S5 of this application constructs a power grid planning optimization model with the goal of minimizing the total lifecycle cost. This model comprehensively considers investment costs (line and substation construction), operating costs (line losses and fuel costs), and maintenance costs, while also incorporating multiple constraints such as power balance, equipment capacity, renewable energy integration (integration rate not less than 90%), and entity behavior. The optimal planning scheme for each scenario is evaluated through weighted scoring (combining renewable energy integration rate, N-1 safety pass rate, and average annual cost). The final determined planning scheme controls the total lifecycle cost while ensuring the safe and stable operation of the system and a high proportion of renewable energy integration, achieving a multi-objective optimization balance and solving the performance shortcomings of existing planning schemes driven by a single objective. Attached Figure Description

[0021] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0022] Figure 1 This is a flowchart illustrating a method for constructing a power grid planning scenario based on the simulation of the main behaviors of power system entities. Figure 2 This is a schematic diagram of the structure of a power grid planning scenario based on the simulation of the main behavior of the power system. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] It should be understood that, when used in this application, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof.

[0025] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0026] It should also be further understood that the term “and / or” as used in this application refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes such combinations.

[0027] Example 1 like Figure 1 As shown, a method for constructing a power grid planning scenario based on the simulation of the main behaviors of the power system includes the following steps: S1: Simulate and analyze the behavior of the main components of the power system to obtain the output and load constraints of each component.

[0028] S11: Subject classification and behavioral boundary definition.

[0029] 1. Classification of core entity types: including thermal power generation enterprises, new energy power generation enterprises (photovoltaic / wind power), power grid operation enterprises, industrial users, and residential users.

[0030] 2. Clarify the constraints for each entity: Power generation companies: must meet the upper and lower limits of unit output (e.g., minimum stable output of thermal power). , maximum output Constraints on the randomness of new energy output (affected by irradiance and wind speed) and constraints on unit start-up and shutdown time (such as minimum start-up and shutdown intervals for thermal power plants). ).

[0031] Power grid operators: Must meet the upper limit of line transmission power Substation capacity constraints N-1 safety principle (the system continues to operate stably after a single component failure).

[0032] Users: Must meet the rigid range of basic electricity demand (such as the minimum guaranteed production load for industrial users). Basic value of residential users' living load ).

[0033] S12: Quantify the behaviors and constraints of each entity, specifically including: 1. Thermal power plants (output optimization model) Objective function (minimize the output allocation corresponding to fuel consumption cost): ; in, The unit output of the thermal power plant at time t (unit: MW); a, b, and c are the fuel consumption coefficients of the thermal power unit (calibrated based on unit design parameters and measured data, units: t / (MW²·h), t / (MW·h), t / h).

[0034] Constraints: (Output upper and lower limit constraints); (Unit ramp rate constraint) For the maximum load increase rate, (Maximum load reduction rate).

[0035] 2. New energy power generation enterprises (output probability model): Photovoltaic power output calculation: ; in, Rated power of photovoltaic power station (unit: MW); for Actual solar irradiance at any given time (unit: W / m²) ; Irradiance under standard test conditions (value taken as 1000 W / m²) ; The conversion efficiency of the photovoltaic module (determined based on the module type, typically 0.18~0.25).

[0036] Wind power output: ; in, The cut-in wind speed of the fan (unit: m / s, usually 3~4 m / s). Rated wind speed of the fan (unit: m / s, usually 12~15m / s). The cutoff velocity for the fan (unit: m / s, usually 25~30m / s). The actual wind speed at time t (unit: m / s). Rated power of the wind turbine (unit: MW).

[0037] 3. User-side demand response behavior: Adjustable load for industrial users: ; in, for Adjustable load capacity for industrial users at any time (unit: MW); The load adjustment factor for industrial users (determined based on the flexibility of production processes, with a value ranging from 0.3 to 0.8); Maximum load for industrial users (unit: MW); for Total system load at any time (unit: MW); The system's average daily load (unit: MW).

[0038] Adjustable load for residential users: ; in, for Adjustable load capacity for residential users at any time (unit: MW); The load adjustment factor for residential users (determined based on energy consumption habits, with a value ranging from 0.1 to 0.4); for Outdoor temperature at any time (unit: °C); The average of the comfortable temperature range for residents (unit: °C, usually 22). C).

[0039] 4. Power Grid Operating Companies (Operation, Maintenance, and Dispatch Model) Line loss calculation: ; in, for Line power loss at any time (unit: MW); for Line transmission power at any time (unit: MW); Line resistance (unit: (Calculation based on conductor material and length) This refers to the line's rated voltage (unit: kV).

[0040] S13: Based on system power balance constraints, a multi-agent behavior equilibrium model is constructed. The stable operating points of each agent are solved using the Newton-Raphson method. After solving, the set of stable behavior parameters of each agent under different operating conditions is output. The formula for the multi-agent behavior equilibrium model is as follows: ; in, For the first Power generation entities Power output at all times (unit: MW); For the first individual users Load at any time (unit: MW); For the first Line Time-of-use loss (unit: MW); , , These are the power generation entities, users, and the number of power lines, respectively.

[0041] S2: Extract representative basic operating modes from historical power system operating data, reduce scenario redundancy through clustering algorithms, and form a basic scenario set covering normal operating states, providing underlying support for subsequent scenario expansion.

[0042] S21: Select the core feature vectors for clustering and perform preprocessing.

[0043] 1. Based on key indicators of power system operation, the feature vector is determined as follows: ; ; ; in, for Renewable energy penetration rate at any given time (no unit). The total output of all renewable energy sources (photovoltaic and wind power) at time t (unit: MW); The total output of all power generation entities (thermal power, new energy, etc.) in the system at time t (unit: MW); for Total system load at any time (unit: MW); for Peak-to-valley load difference rate at any time (unitless). The maximum load (in MW) within the statistical period at time t. The minimum load (unit: MW) within the statistical period at time t. The average load (unit: MW) during the statistical period at time t. for Total power loss of the system at any time (unit: MW).

[0044] 2. Standardize the feature vectors using Min-Max to eliminate the influence of dimensions. The formula is as follows: ; in, , These are the historical minimum and maximum values ​​of the feature variable, respectively, after standardization. .

[0045] S22: The K-means algorithm is used to cluster the core feature variables.

[0046] 1. Determine the number of clusters The elbow method and silhouette coefficient are used to jointly determine and calculate different... The sum of squared errors (SSE) within clusters corresponding to values ​​(usually 3-6) and the silhouette coefficient. Select When the maximum SSE decreases at a slowdown Value (typical) (This corresponds to high / medium / low penetration rate-load combination scenarios).

[0047] 2. Clustering objective function: Minimize the sum of squared Euclidean distances between the feature vectors within each cluster, as shown in the formula: min ; in, For the first Each cluster (corresponding to a basic scenario); For the first The center vector of each cluster (after standardization); The distance is Euclidean.

[0048] S23: Calculate the mean of the feature vector for each cluster, and restore it to the actual physical quantity (de-standardization) to obtain the key parameters of each basic scenario, forming a set of basic scenarios. .For example: High penetration rate - high load scenario ( ); Medium penetration rate - medium load scenario ( ); Low penetration rate - low load scenario ( ); Medium penetration rate - high load scenario ( ).

[0049] S3: Using the LSTM sequence prediction model, with historical power system operation data as input, it predicts the system operation status for different time periods in the future and generates future operation prediction data.

[0050] Based on historical power system operation data and environmental data (such as irradiance, wind speed, and temperature), the LSTM time series prediction model accurately predicts the system operation status in different time periods (short-term 24h, medium-term 72h, and long-term 168h), providing data support for extending the basic scenario to future operating conditions.

[0051] S31: Determine the input feature sequence and output prediction vector of the prediction model.

[0052] Input feature sequence: Select 168 hours (7 days) of historical operational and environmental data to construct the input vector. ,in Irradiance (unit: W / m²) , Wind speed (unit: m / s) Temperature (unit: °C).

[0053] Output prediction vector: the system's operating state for the future target time period, i.e. ( For predicted duration, in hours.

[0054] S32: Construct an LSTM sequence prediction model and train it.

[0055] Model architecture: Input layer (number of neurons = input feature dimension) LSTM hidden layer (2 layers, 128 neurons per layer, activation function is tanh) → Dropout layer (dropout rate=0.2, to prevent overfitting) → Fully connected layer (number of neurons = output feature dimension) → Output layer (activation function is linear).

[0056] Loss function: Root mean square error (RMSE) is used, and the formula is as follows: ; in, For the sample size, To output feature dimensions, For the first The first sample The predicted value of each feature, This corresponds to the actual value.

[0057] Training parameters: Optimizer is Adam, learning rate = 0.001, batch size = 32, number of epochs = 100, early stopping is used to prevent overfitting (patience = 10).

[0058] S33: Correct the prediction results of the LSTM sequence prediction model to obtain future prediction data.

[0059] An environmental factor correction coefficient is introduced to reduce prediction bias; the formula is as follows: ; in, This is the revised forecast data for future operation; Environmental impact coefficients (obtained through regression analysis of historical environmental and operational data, such as the impact coefficient of wind speed on wind power output). ; for The deviation rate between the predicted environmental data and the historical average environmental data at any given time ( (No unit).

[0060] S4: Combining the basic operating scenarios obtained from clustering with the predicted future operating data, select key scenarios covering high, medium, and low renewable energy penetration rates, high, medium, and low load levels, and the impact of extreme weather (such as typhoons and cold waves) to construct a typical scenario library for power grid planning.

[0061] From the combination of "basic scenarios + future forecast data", key scenarios covering high / medium / low renewable energy penetration rates, high / medium / low load levels and extreme weather conditions are selected, redundant scenarios are eliminated, and a typical scenario library that can support power grid planning is constructed.

[0062] S41: Generate combined scenarios based on basic scenarios and future operational prediction data.

[0063] Based on the basic scene set With future forecast data Generate composite scenes using Cartesian products : It covers a combination of operating conditions with different penetration rates, load levels, and time dimensions (total). (A basic combination, divided into time dimensions by day).

[0064] S42: Set the filtering criteria for selecting combined scenarios.

[0065] Four categories of screening criteria were set to ensure both scenario coverage and differentiation: Indicator 1: Renewable Energy Penetration Coverage Must include high ,middle ,Low There are three intervals, and each interval must retain at least one scene.

[0066] Indicator 2: Load Level Coverage Must include high ),middle ),Low Three intervals, each interval At least one scene must be reserved.

[0067] Indicator 3: Coverage of extreme operating conditions Extreme weather scenarios must meet the following requirements: Typhoon conditions: (Wind speed exceeds historical average by 50%), and wind power output fluctuates. ; Cold wave conditions: (Temperature is 30% below the historical average), and residential heating load has increased. ; At least one scenario should be retained for each extreme operating condition.

[0068] Indicator 4: Scenario Differences The Euclidean distance (after standardization) between the feature vectors of any two scenes must satisfy the following condition: To avoid scene repetition (the smaller the distance, the higher the scene similarity).

[0069] S43: After filtering based on screening criteria, retain key scenarios to form a typical scenario library.

[0070] After screening, retain 8-12 key scenarios to form a typical scenario library: ( ; For example: Typical scenarios (6): High permeability - High load, High permeability - Medium load, Medium permeability - High load, Medium permeability - Low load, Low permeability - High load, Low permeability - Low load; Extreme scenarios (2-4): Typhoon - high permeability, cold wave - high load, extreme low wind speed - low permeability, etc.

[0071] S5: Using scenarios from the typical scenario library as input, import them into the power grid planning optimization model, combine the simulation results of the main behavior of the power system to conduct simulation evaluation, and determine the final power grid planning scheme.

[0072] Using key scenarios from the typical scenario library as input, and combining the simulation results of the main behaviors output by S1, the power grid planning optimization model is used to simulate and evaluate planning schemes (such as new lines, substation expansion, and new energy grid connection) to verify the feasibility and optimality of the schemes.

[0073] S51: Construct a power grid planning optimization model.

[0074] 1. The objective function (minimum life-cycle cost) of the power grid planning optimization model is as follows: ; ; ; ; in, For investment costs; Cost of line per unit cross-section (unit: RMB 10,000 / (mm²·km)); Line length (unit: km); The cross-sectional area of ​​the line (unit: mm²); Cost per unit capacity of substation (unit: RMB 10,000 / MVA). Substation capacity (unit: MVA); Operating costs; The unit cost of electricity is the fuel cost (unit: yuan / MWh), and 8760 represents the number of hours per year. Maintenance costs (annual maintenance fee calculated at 2% of investment cost, based on power industry operation and maintenance experience).

[0075] 2. Constraints of the power grid planning optimization model: Power balance constraints: (With S1 equilibrium solution as constraint); Equipment capacity constraints (Line transmission limit) (Substation capacity limit); Constraints on the absorption of new energy sources: (The renewable energy consumption rate shall not be less than 90%); Subject Behavior Constraints: The output and load constraints of each subject in S1 are adopted.

[0076] S52: Import each scenario into the power grid planning optimization model to obtain the corresponding optimal planning scheme, and determine the evaluation index value of the optimal planning scheme for each scenario.

[0077] The runtime parameters for each scenario in the typical scenario library ( , The main behavioral parameters are imported into the power grid planning optimization model, and an improved genetic algorithm (introducing an elite retention strategy) is used to solve the model, outputting the optimal planning scheme under each scenario (such as adding 3 new 220kV lines and expanding 2 110kV substations).

[0078] The evaluation metrics for the optimal planning scheme in each scenario are determined as follows: New energy consumption rate: N-1 pass rate: Average annual cost: ( (The lifespan of the power grid equipment is taken as 20 years).

[0079] S53: Obtain the corresponding weighted score based on the evaluation index value of the optimal planning scheme in each scenario, and determine the final power grid planning scheme based on the weighted score.

[0080] The Analytic Hierarchy Process (AHP) is used to determine the weights of the indicators (e.g., , , ), calculate the weighted score for each option: in, , These represent the minimum and maximum annual average costs for each scheme (cost indicators are normalized and converted into positive scores). The scheme with the highest score is selected as the final power grid planning scheme.

[0081] The beneficial effects of this invention are as follows: 1. Improve the adaptability of scene construction to actual working conditions and solve the problem of incomplete characterization of main behaviors. Existing technologies often rely on experience or simple data statistics to set planning scenarios, failing to consider the differences in behavior among multiple stakeholders, such as the output optimization of thermal power plants, the stochastic generation of renewable energy plants, and user demand response. This leads to a disconnect between the scenario and reality. This application simulates and analyzes the behavior of the main stakeholders in the power system through step S1, clarifying the output and load constraints of each stakeholder (such as the upper and lower limits of thermal power plant unit output, the stochastic constraints of renewable energy output, and the rigid range of user electricity demand). It deeply integrates the behavioral characteristics of stakeholders into the entire scenario construction process, enabling the planned scenario to accurately reflect the decision-making logic and operational rules of multiple stakeholders in the actual system. This significantly improves the fit between the scenario and actual operating conditions, laying the foundation for the practicality of subsequent planning schemes.

[0082] 2. Improve the accuracy of future operating condition predictions, covering multi-dimensional combinations of operating conditions. Traditional methods lack scientific time-series models to predict future operating conditions and do not incorporate environmental factor corrections, making it difficult to cover different renewable energy penetration rates, load levels, and extreme weather conditions. Step S3 of this application employs an LSTM sequence prediction model, using historical operating data as input, and incorporates environmental impact coefficients (such as the impact of wind speed and temperature on renewable energy output) to correct the prediction results, accurately predicting the system's operating status at different future time periods. Step S4 further combines basic scenarios with corrected future prediction data to select key scenarios covering high, medium, and low renewable energy penetration rates, high / medium / low load levels, and the impact of extreme weather (typhoons, cold waves), filling the gap in multi-dimensional operating condition coverage in existing technologies and ensuring that the planning scheme can cope with complex and ever-changing future operating environments.

[0083] 3. Optimize scene quality and improve planning simulation efficiency. Existing technologies suffer from high redundancy in scenario construction and insufficient extraction of key scenarios, leading to large computational loads and low efficiency in planning simulation. This application addresses this by using the K-means clustering algorithm in step S2 to cluster basic operating patterns extracted from historical operating data, reducing scenario redundancy and forming a basic scenario set covering normal operating states. Step S4 further filters key scenarios based on clear selection criteria (renewable energy penetration coverage, load level coverage, extreme operating condition coverage, and scenario differences), constructing a concise yet comprehensive library of typical scenarios. This significantly reduces the interference of invalid scenarios on simulation calculations, improves the computational efficiency of the power grid planning optimization model, and ensures the representativeness and coverage of scenarios, avoiding the waste of planning resources due to scenario redundancy.

[0084] 4. Achieve multi-objective optimization of the planning scheme, balancing cost and operational performance. Traditional planning schemes often struggle to balance multiple objectives, including investment costs, operational efficiency, and renewable energy integration. Step S5 of this application constructs a power grid planning optimization model with the goal of minimizing the total lifecycle cost. This model comprehensively considers investment costs (line and substation construction), operating costs (line losses and fuel costs), and maintenance costs, while also incorporating multiple constraints such as power balance, equipment capacity, renewable energy integration (integration rate not less than 90%), and entity behavior. The optimal planning scheme for each scenario is evaluated through weighted scoring (combining renewable energy integration rate, N-1 safety pass rate, and average annual cost). The final determined planning scheme controls the total lifecycle cost while ensuring the safe and stable operation of the system and a high proportion of renewable energy integration, achieving a multi-objective optimization balance and solving the performance shortcomings of existing planning schemes driven by a single objective.

[0085] Example 2 like Figure 2 As shown, a power grid planning scenario construction system based on power system agent behavior simulation is applied to the aforementioned power grid planning scenario construction method based on power system agent behavior simulation, and includes: The analysis module is used to simulate and analyze the behavior of the main components of the power system, and obtain the output and load constraints of each component. The clustering module is used to extract basic operating patterns from historical operating data of the power system, reduce scenario redundancy through clustering algorithms, form a basic scenario set covering the normal operating state, and provide underlying support for subsequent scenario expansion. The prediction module is used to use an LSTM sequence prediction model, taking historical power system operation data as input, to predict the system operation status for different time periods in the future and generate future operation prediction data. The scenario module is used to combine basic operation scenarios with future operation forecast data to select key scenarios covering high, medium and low renewable energy penetration rates, high, medium and low load levels, and the impact of extreme weather, and to build a typical scenario library for power grid planning. The simulation module is used to take scenarios from the typical scenario library as input, import them into the power grid planning and optimization model, combine the simulation results of the main behavior of the power system to conduct simulation evaluation, and determine the final power grid planning scheme.

[0086] Example 3 A computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to execute the above-described method for constructing a power grid planning scenario based on simulation of the main behavior of a power system.

[0087] Example 4 A processor for running a program, wherein the program executes the above-described method for constructing a power grid planning scenario based on simulation of the main behavior of a power system.

[0088] This application discloses a method, system, medium, and processor for constructing power grid planning scenarios based on the simulation of power system entity behavior, relating to the field of power system planning technology. It first simulates and analyzes the behavior of power system entities to obtain the output and load constraints of each entity; then, it extracts basic operating modes from historical operating data and forms a basic scenario set using a clustering algorithm; subsequently, it uses an LSTM sequence prediction model combined with environmental correction to generate future operating prediction data; next, it combines the basic scenarios and future data to screen key scenarios covering different renewable energy penetration rates, load levels, and extreme weather conditions, constructing a typical scenario library; finally, it imports the typical scenarios into a power grid planning optimization model, combines the simulation results of entity behavior, and determines the final planning scheme. This invention improves the adaptability of scenarios to actual operating conditions, ensuring that the planning scheme takes into account cost, safety, and renewable energy consumption, and is suitable for precise power grid planning.

[0089] Those skilled in the art will recognize that the units of the various examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of the invention.

[0090] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical functional division. In actual implementation, there may be other division methods, such as multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored.

[0091] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0092] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0093] 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 them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of this application.

Claims

1. A method for constructing power grid planning scenarios based on simulation of power system entity behavior, characterized in that, include: S1: Simulate and analyze the behavior of the main components of the power system to obtain the output and load constraints of each component; S2: Extract basic operating modes from historical power system operating data, reduce scenario redundancy through clustering algorithms, form a basic scenario set covering normal operating states, and provide underlying support for subsequent scenario expansion; S3: Using the LSTM sequence prediction model, with historical power system operation data as input, it predicts the system operation status for different time periods in the future and generates future operation prediction data; S4: Combining basic operation scenarios and future operation forecast data, select key scenarios covering high, medium and low renewable energy penetration rates, high, medium and low load levels, and the impact of extreme weather, and construct a typical scenario library for power grid planning; S5: Using scenarios from the typical scenario library as input, import them into the power grid planning optimization model, combine the simulation results of the main behavior of the power system to conduct simulation evaluation, and determine the final power grid planning scheme.

2. The method for constructing a power grid planning scenario based on power system subject behavior simulation according to claim 1, characterized in that, In step S2, the extraction of basic operating modes from historical power system operating data, and the reduction of scenario redundancy through clustering algorithms to form a basic scenario set covering normal operating states, includes the following steps: S21: Select the core feature vectors for clustering and perform preprocessing; S22: The K-means algorithm is used to cluster the core clustering feature variables; S23: Calculate the mean of the feature vectors for each cluster, convert them back to actual physical quantities, obtain the key parameters of each basic scenario, and form a set of basic scenarios. , Based on scenario K.

3. The method for constructing a power grid planning scenario based on the simulation of power system subject behavior according to claim 1, characterized in that, In step S3, the LSTM sequence prediction model, using historical power system operating data as input, predicts the system operating status for different future time periods to generate future operating prediction data, including the following steps: S31: Determine the input feature sequence and output prediction vector of the prediction model; S32: Construct an LSTM sequence prediction model and train it; S33: Correct the prediction results of the LSTM sequence prediction model to obtain future prediction data; the formula for correcting the prediction results is as follows: ； in, This is the revised forecast data for future operation; This is the environmental impact coefficient; for The deviation rate between real-time predicted environmental data and historical average environmental data; This is the output prediction vector of the LSTM sequence prediction model.

4. The method for constructing a power grid planning scenario based on the simulation of power system subject behavior according to claim 3, characterized in that, In step S4, the process of combining basic operational scenarios with future operational forecast data to select key scenarios covering high, medium, and low renewable energy penetration rates, high, medium, and low load levels, and the impact of extreme weather, and constructing a typical scenario library for power grid planning, includes the following steps: S41: Generate a combined scenario based on the basic scenario and future operational prediction data, using the following formula: ； In the formula, A collection of basic scenarios; These are revised future forecasts; For combined scenes; S42: Set screening criteria for screening combined scenarios, including renewable energy penetration coverage, load level coverage, extreme operating condition coverage, and scenario differences; S43: After filtering based on the screening criteria, key scenarios are retained to form a typical scenario library as follows: , Scenes that have been selected and retained.

5. The method for constructing a power grid planning scenario based on power system subject behavior simulation according to claim 4, characterized in that, In step S5, the process of using scenarios from the typical scenario library as input, importing them into the power grid planning optimization model, and combining the simulation results of the main behaviors of the power system for simulation evaluation to determine the final power grid planning scheme includes the following steps: S51: Construct a power grid planning optimization model; S52: Import each scenario into the power grid planning optimization model to obtain the corresponding optimal planning scheme, and determine the evaluation index value of the optimal planning scheme for each scenario; S53: Obtain the corresponding weighted score based on the evaluation index value of the optimal planning scheme in each scenario, and determine the final power grid planning scheme based on the weighted score.

6. The method for constructing a power grid planning scenario based on power system subject behavior simulation according to claim 5, characterized in that, The objective function of the power grid planning optimization model is as follows: ； ； ； ； in, For investment costs; Cost per unit cross-section of the circuit; This refers to the line length; The cross-sectional area of ​​the line; The cost per unit capacity of a substation. For substation capacity; Operating costs; The cost of fuel per unit of electricity; To maintain costs; The constraints of the power grid planning optimization model are as follows: Power balance constraints: ; Equipment capacity constraints , ; Constraints on the absorption of new energy sources: ; Constraints on the behavior of the main entities: the output and load constraints of each entity; In the above formula, This is the upper limit for line transmission; This is the upper limit of substation capacity; For the first Power generation entities Always put in the effort; For the first individual users Constant load; For the first Line Time loss; , , These are the power generation entities, users, and the number of power lines, respectively. Let t be the actual amount of renewable energy power generated by the system at time t; Let t be the total power generation of the new energy power generation enterprise.

7. The method for constructing a power grid planning scenario based on power system subject behavior simulation according to claim 6, characterized in that, In step S53, the formula for determining the weighted score is as follows: ； in, For weighted scores; , These represent the minimum and maximum annual average costs for each option; , , These are the weights of the corresponding indicators; This represents the average annual cost. For the renewable energy consumption rate; The safe pass rate is N-1.

8. A power grid planning scenario construction system based on power system entity behavior simulation, characterized in that, The method for constructing a power grid planning scenario based on power system agent behavior simulation as described in any one of claims 1 to 7 includes: The analysis module is used to simulate and analyze the behavior of the main components of the power system, and obtain the output and load constraints of each component. The clustering module is used to extract basic operating patterns from historical operating data of the power system, reduce scenario redundancy through clustering algorithms, form a basic scenario set covering the normal operating state, and provide underlying support for subsequent scenario expansion. The prediction module is used to use an LSTM sequence prediction model, taking historical power system operation data as input, to predict the system operation status for different time periods in the future and generate future operation prediction data. The scenario module is used to combine basic operation scenarios with future operation forecast data to select key scenarios covering high, medium and low renewable energy penetration rates, high, medium and low load levels, and the impact of extreme weather, and to build a typical scenario library for power grid planning. The simulation module is used to take scenarios from the typical scenario library as input, import them into the power grid planning and optimization model, combine the simulation results of the main behavior of the power system to conduct simulation evaluation, and determine the final power grid planning scheme.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to execute the power grid planning scenario construction method based on power system subject behavior simulation as described in any one of claims 1 to 7.

10. A processor, characterized in that, The processor is used to run a program, wherein the program executes the power grid planning scenario construction method based on power system subject behavior simulation as described in any one of claims 1 to 7.

Images