A method and system for dynamic regulation of regional energy storage based on simulation model

Abstract

This invention proposes a regional energy storage dynamic control method and system based on simulation model. The method includes: constructing a simulation model based on historical data of equipment technical parameters, load history and real-time data, energy storage system specifications history, distributed photovoltaic power generation history and real-time data, and weather history and real-time data for each distribution area; based on the simulation model, monitoring in real time whether each distribution area experiences reverse overload and whether the distribution area is currently in the peak period of distributed photovoltaic power generation; according to the trigger condition type and the energy storage status of each distribution area, using particle swarm optimization algorithm to generate the optimal charging strategy for the energy storage equipment; and performing dynamic energy storage scheduling for the energy storage equipment in each distribution area according to the generated optimal charging strategy. By combining the accurate simulation of the simulation model with the efficient search capability of the particle swarm optimization algorithm, the control cost is effectively saved and the reliability of energy storage control is improved.

Description

Technical Field

[0001] This invention relates to the field of power distribution network operation technology, and in particular to a method and system for dynamic regulation of regional energy storage based on simulation models. Background Technology

[0002] With the large-scale and high-proportion integration of new energy sources such as wind power and photovoltaics into the power grid, their inherent intermittency and volatility pose new challenges to the stability of the power system. Energy storage technology, as an effective means of mitigating the fluctuations of these renewable energy sources, is crucial for maintaining grid stability and improving operational efficiency. However, currently, the dispatch and operation mechanisms for new energy storage systems are incomplete, resulting in phenomena such as "built but not used" and "built but not regulated," leading to low utilization rates. Therefore, developing a new type of dynamic regulation and control system for energy storage is essential.

[0003] In related technologies, dynamic regulation of energy storage is generally based on setting regulation thresholds according to the line operation or the time period, without considering reverse overload and photovoltaic power generation factors, nor does it consider monitoring the grid operation through simulation models, resulting in low reliability and economy of energy storage regulation.

[0004] To address this problem, the present invention provides a method and system for dynamic regulation of regional energy storage based on simulation models, in order to solve the aforementioned issues. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention innovatively proposes a regional energy storage dynamic control method and system based on simulation model, which effectively solves the problems of low reliability and economy of energy storage control caused by the prior art, effectively saves control costs, and improves the reliability of energy storage control.

[0006] The first aspect of this invention provides a method for dynamic regulation of regional energy storage based on simulation models, comprising:

[0007] Acquire historical data on equipment technical parameters, load, and real-time data for each transformer substation, as well as historical and real-time data on energy storage system specifications, distributed photovoltaic power generation, and weather.

[0008] Based on historical data of equipment technical parameters, load, and real-time data in each transformer substation, historical data of energy storage system specifications, historical data of distributed photovoltaic power generation, and historical data of weather, a simulation model is constructed.

[0009] Based on the simulation model, the system monitors in real time whether each transformer area experiences reverse overload and whether it is currently in the peak period of distributed photovoltaic power generation. If the current transformer area meets the preset algorithm scheduling trigger conditions, the optimal charging strategy for the energy storage device is generated using the particle swarm optimization algorithm according to the trigger condition type and the energy storage status of each transformer area. The preset algorithm scheduling trigger conditions include a first trigger condition and a second trigger condition. The first trigger condition is that a first preset number of transformer areas in the region all experience reverse overload, and the first preset number of transformer areas in the region are all in the peak period of distributed photovoltaic power generation. The second trigger condition is that the sum of the number of transformer areas in the region and the number of transformer areas in the peak period of distributed photovoltaic power generation is a second preset number, wherein the first preset number is less than the second preset number.

[0010] Based on the generated optimal charging strategy for energy storage devices, dynamic energy storage scheduling is performed on the energy storage devices in each distribution area to eliminate reverse overload in the distribution area.

[0011] Optionally, based on historical data of equipment technical parameters, historical and real-time load data, historical data of energy storage system specifications, historical and real-time data of distributed photovoltaic power generation, and historical and real-time weather data within each transformer substation area, a simulation model is constructed, including:

[0012] Based on historical data of equipment technical parameters and energy storage system specifications within each distribution area, a power grid model and an equipment electrical model are established. The power grid model includes a topology model of the power grid within different distribution areas, used to represent the connection relationships between various equipment nodes within different distribution areas, displaying the topology and electrical characteristics of the distribution network, providing an operating environment for other models, performing power flow calculations based on data predicted by the distributed photovoltaic power generation prediction model, and feeding the results back to the equipment model and load model. The equipment electrical model includes a distribution transformer model, a circuit breaker model, and a distributed power source model, used to represent the electrical parameters of each device within different distribution areas, determining the role of the device in the power system based on the location information in the power grid model, displaying relevant electrical data of the device, and adjusting the status or output of the device based on the calculation results of the power grid model.

[0013] Based on historical and real-time load data, as well as historical and real-time weather data for each distribution area, a load model based on an LSTM neural network is constructed. This model is used to predict the load curves of each distribution area based on the historical and real-time load data, as well as the historical and real-time weather data. The load curves are then adjusted according to the calculation results of the power grid model to adapt to different operating conditions.

[0014] Based on historical and real-time data of distributed photovoltaic power generation in each transformer substation, as well as historical and real-time weather data, a distributed photovoltaic power generation prediction model based on LSTM neural network is constructed. This model is used to predict the distributed photovoltaic power generation of each transformer substation based on the historical and real-time data of distributed photovoltaic power generation and the historical and real-time weather data, and the prediction results are transmitted to the power grid model in the form of a file.

[0015] Furthermore, based on simulation models, real-time monitoring is conducted to determine whether each distribution area experiences reverse overload and whether the distribution area is currently in the peak period of distributed photovoltaic power generation.

[0016] When current flows from the user to the power grid in a distribution area, and the transformer load rate is higher than a first preset value, then the distribution area experiences a reverse overload; the transformer load rate is calculated as follows:

[0017]

[0018] Where, β 反向 For transformer load factor; S 反向 It is the reverse apparent power of the total meter for the distribution area; S e This is the rated capacity of the transformer;

[0019] When a distribution area is in a period of expected sufficient sunlight, and the rate of change of the power generation curve predicted by the corresponding distributed photovoltaic power generation prediction model for that area exceeds a preset rate of change threshold during the current period of expected sufficient sunlight, then the distribution area is currently in the peak period of distributed photovoltaic power generation.

[0020] Optionally, based on the triggering condition type and the energy storage status of each distribution area, the optimal charging strategy for the energy storage device is generated using a particle swarm optimization algorithm, specifically including:

[0021] Based on the triggering condition type and the energy storage situation of each distribution area, the objective function and constraints of regulation are determined. The objective function is the sum of the grid overload and the charging and discharging cost of the energy storage device, and the constraint is that no reverse heavy overload occurs in each distribution area.

[0022] A particle swarm optimization algorithm with added constraints is introduced to generate the optimal charging strategy for energy storage devices.

[0023] Furthermore, when the trigger condition type is the first trigger condition, the objective function is:

[0024]

[0025] Where Objective 1 is the minimum sum of grid overload and energy storage device charging / discharging costs under the first triggering condition; N1 is the number of distribution substations experiencing reverse overload and during peak distributed photovoltaic (PV) generation; Q1 is the number of distribution substations not experiencing reverse overload and not during peak distributed PV generation; M is the number of energy storage devices within the distribution substation; PV excess,i It refers to the excess photovoltaic power generation of transformer area i, which is experiencing reverse overload and is in the peak period of distributed photovoltaic power generation. Cost is the amount of electricity allocated to energy storage devices j within transformer area α1 that have not experienced reverse overload and are not in the peak period of distributed photovoltaic power generation; Allocate,αj It is the total cost allocated to the energy storage device j in the distribution area α1 that has not experienced reverse heavy overload and is not in the peak period of distributed photovoltaic power generation. The total cost includes charging and discharging loss cost and equipment depreciation cost.

[0026] When the trigger condition type is the second trigger condition, the objective function is:

[0027]

[0028] Where Objective2 is the minimum sum of grid overload and energy storage device charging / discharging cost when the first trigger condition is met; N2 is the number of distribution transformer areas experiencing reverse overload or during peak distributed photovoltaic power generation; Q2 is the number of distribution transformer areas without reverse overload; M is the number of energy storage devices within the distribution transformer area; PV excess,i It refers to the excess photovoltaic power generation of transformer area i, which is experiencing reverse overload and is in the peak period of distributed photovoltaic power generation. It is the amount of electricity allocated to the energy storage device j in the transformer area α2 where no reverse overload has occurred; It is the total cost allocated to energy storage device j in transformer area α2 where no reverse overload has occurred. The total cost includes charging and discharging loss costs and equipment depreciation costs.

[0029] Furthermore, when the triggering condition type is the first triggering condition, the constraints all include grid carrying capacity constraints and energy storage device capacity constraints, wherein the grid carrying capacity constraints are:

[0030]

[0031] Among them, L t TotalLoad represents the grid carrying capacity at time t. t This represents the total load of the power grid at time t;

[0032] The capacity constraints for energy storage devices are:

[0033]

[0034] Among them, C max,jIt is the maximum capacity that the j-th energy storage device can store;

[0035] When the trigger condition type is the second trigger condition, the constraints all include grid carrying capacity constraints and energy storage device capacity constraints. Among them, the grid carrying capacity constraints are:

[0036]

[0037] The capacity constraints for energy storage devices are:

[0038]

[0039] Furthermore, if a violation of the constraints occurs, the objective function is adjusted using a penalty function;

[0040] The penalty function includes the equipment capacity penalty function Penalty. capacity,αj Penalty function for grid carrying capacity load,t When the trigger condition type is the first trigger condition, the device capacity penalty function The expression is:

[0041]

[0042] When the trigger condition type is the second trigger condition, the device capacity penalty function The expression is:

[0043]

[0044] When the trigger condition type is the first trigger condition, the grid carrying capacity penalty function Penalty load,t,1 The expression is:

[0045]

[0046] When the triggering condition type is the second triggering condition, the grid carrying capacity penalty function Penalty load,t,2 The expression is:

[0047]

[0048] When the trigger condition type is the first trigger condition, the adjusted new objective function is:

[0049]

[0050] Where λ is the penalty function coefficient, used to control the influence of the penalty function on the objective function;

[0051] When the trigger condition type is the second trigger condition, the adjusted new objective function is:

[0052]

[0053] Where λ is the penalty function coefficient, used to control the influence of the penalty function on the objective function.

[0054] Optionally, a particle swarm optimization algorithm with added constraints is introduced to generate the optimal charging strategy for the energy storage device, specifically including:

[0055] Initialize the particle position parameters. The position of each particle can be represented as an M×N matrix, where each element... This represents the proportion or absolute value of the electricity allocated to energy storage device j in distribution area α1, or, for each element. This represents the proportion or absolute value of the electricity allocated to energy storage device j in transformer area α2;

[0056] Calculate the fitness value, where the fitness value is the objective function value;

[0057] Determine if any particle violates the constraints. If a particle violates the constraints, apply the penalty function to adjust the objective function, and recalculate the fitness value of each particle based on the adjusted objective function.

[0058] Based on the individual extreme value and global extreme value of each particle, update the particle position and velocity, where the update expression for particle position and velocity is:

[0059] v mn (t+1)=a(t)v mn (t)+b1r1(t)(p mn (t)-y mn (t))+b2r2(t)(p gn (t)-y mn (t)),

[0060] y mn (t+1)=y mn (t)+v mn (t+1),

[0061] Among them, v mn (t), y mn (t), p mn (t) represents the velocity, position, and individual extreme value of the m-th particle in dimension n at the t-th iteration; p gn (t) represents the global extremum in dimension n at the t-th iteration; a(t) represents the magnitude of the inertia weight at the t-th iteration; b1 and b2 are learning factors; r1(t) and r2(t) are random numbers in the interval (0,1].

[0062] The position and velocity of each particle are iteratively updated until a preset stopping condition is met. The preset stopping condition includes reaching the maximum number of iterations or the fitness value being greater than a preset fitness threshold.

[0063] Optionally, before performing dynamic energy storage scheduling on the energy storage devices in each distribution area based on the generated optimal charging strategy for the energy storage devices, the following steps are also included:

[0064] The generated optimal charging strategy for the energy storage device is simulated and verified in the simulation model. If the reverse overload phenomenon disappears in all areas, the optimal charging strategy for the energy storage device is verified. If the reverse overload phenomenon does not disappear in a certain area, the optimal charging strategy for the energy storage device is not verified.

[0065] A second aspect of the present invention provides a regional energy storage dynamic control system based on simulation model, comprising:

[0066] The acquisition module acquires historical data on equipment technical parameters, load, energy storage system specifications, distributed photovoltaic power generation, and weather for each transformer substation.

[0067] The construction module builds a simulation model based on historical data of equipment technical parameters, load historical data and real-time data, energy storage system specifications historical data, distributed photovoltaic power generation historical data and real-time data, and weather historical data and real-time data in each transformer area.

[0068] The monitoring and generation module, based on simulation models, monitors in real time whether each transformer area experiences reverse overload and whether the area is currently in the peak period of distributed photovoltaic power generation. If the current transformer area meets the preset algorithm scheduling trigger conditions, it generates the optimal charging strategy for the energy storage device using the particle swarm optimization algorithm, based on the trigger condition type and the energy storage status of each transformer area. The preset algorithm scheduling trigger conditions include a first trigger condition and a second trigger condition. The first trigger condition is that a first preset number of transformer areas in the region all experience reverse overload, and the first preset number of transformer areas in the region are all in the peak period of distributed photovoltaic power generation. The second trigger condition is that the sum of the number of transformer areas in the region and the number of transformer areas in the peak period of distributed photovoltaic power generation is a second preset number, wherein the first preset number is less than the second preset number.

[0069] The scheduling module performs dynamic energy storage scheduling for each distribution area based on the generated optimal charging strategy for the energy storage devices, in order to eliminate reverse overload in the distribution area.

[0070] The technical solution adopted in this invention has the following technical effects:

[0071] 1. This invention constructs a simulation model based on historical data of equipment technical parameters, load history and real-time data, energy storage system specifications, distributed photovoltaic power generation history and real-time data, and weather history and real-time data for each power distribution area. Based on the simulation model, it monitors in real time whether each power distribution area experiences reverse overload and whether the area is currently in a peak period of distributed photovoltaic power generation. According to the triggering condition type and the energy storage status of each power distribution area, it uses a particle swarm optimization algorithm to generate the optimal charging strategy for the energy storage equipment. Based on the generated optimal charging strategy, it performs dynamic energy storage scheduling for the energy storage equipment in each power distribution area to eliminate reverse overload. By combining the accurate simulation of the simulation model with the efficient search capability of the particle swarm optimization algorithm, it achieves precise control and optimized scheduling of regional power grid energy storage equipment, effectively solving the problems of low reliability and economy of energy storage regulation caused by existing technologies, effectively saving regulation costs and improving the reliability of energy storage regulation.

[0072] 2. The simulation model in the technical solution of this invention includes a power grid model, an equipment electrical model, a load model, and a distributed photovoltaic power generation prediction model. These models are interconnected and work together to respond to changes in power grid load in real time, intelligently allocate and adjust energy storage resources, effectively alleviate the peak-valley difference problem of the power grid, and improve the operating efficiency and economy of the power grid.

[0073] 3. The technical solution of this invention is based on real-time monitoring using simulation models. When the current in the distribution area flows from the user to the power grid, and the transformer load rate is higher than a first preset value, the distribution area experiences reverse overload. This not only considers the power grid model in the simulation model but also combines the electrical parameters in the equipment electrical model, thus organically combining efficiency and accuracy. When the distribution area is in a period of expected sufficient sunlight, and the rate of change of the power generation curve predicted by the distributed photovoltaic power generation prediction model corresponding to the distribution area exceeds a preset rate of change threshold during the current period of expected sufficient sunlight, the distribution area is currently in the peak period of distributed photovoltaic power generation. This not only considers the time period of the distribution area in the simulation model but also combines the output prediction results of the distributed photovoltaic power generation prediction model, thus organically combining efficiency and accuracy.

[0074] 4. In this invention, the objective function and constraints for regulation are determined based on the triggering condition type and the energy storage status of each distribution area. The objective function is the sum of the grid overload and the charging and discharging costs of the energy storage device under different triggering conditions. The constraint is that no reverse heavy overload occurs in any distribution area. A particle swarm optimization algorithm with added constraints is introduced to generate the optimal charging strategy for the energy storage device. If the constraint is violated, the objective function is adjusted through a penalty function. Not only does the objective function vary depending on the triggering condition type, but the penalty function is also adjusted when the constraint is violated, improving the adaptability of energy storage regulation. Furthermore, by combining the accurate simulation of the simulation model with the efficient search capability of the particle swarm optimization algorithm, precise control and optimized scheduling of regional power grid energy storage devices are achieved. This enhances the adaptability to renewable energy fluctuations and the grid's anti-interference capability, reduces energy waste caused by supply and demand imbalances, and provides an innovative solution for the intelligent management and sustainable development of the power grid.

[0075] 5. Before dynamically scheduling energy storage for each distribution area based on the generated optimal charging strategy for the energy storage device, the technical solution of this invention further includes: simulating and verifying the generated optimal charging strategy for the energy storage device in a simulation model. If the reverse overload phenomenon disappears in each distribution area, the optimal charging strategy for the energy storage device is verified successfully; if the reverse overload phenomenon does not disappear in a certain distribution area, the optimal charging strategy for the energy storage device fails to be verified. This ensures the effectiveness of the implementation of the optimal charging strategy for the energy storage device and improves the economic efficiency of energy storage regulation.

[0076] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Attached Figure Description

[0077] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0078] Figure 1 This is a flowchart illustrating the method of Embodiment 1 in the present invention;

[0079] Figure 2 This is a schematic diagram illustrating the implementation process of the method in Embodiment 1 of the present invention;

[0080] Figure 3 This is another flowchart illustrating the method of Embodiment 1 in the present invention;

[0081] Figure 4 This is a schematic diagram of the system structure in Embodiment 2 of the present invention. Detailed Implementation

[0082] To clearly illustrate the technical features of this solution, the invention will be described in detail below through specific embodiments and in conjunction with the accompanying drawings. The following disclosure provides many different embodiments or examples for implementing different structures of the invention. To simplify the disclosure of the invention, components and arrangements of specific examples are described below. Furthermore, reference numerals and / or letters may be repeated in different examples. This repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed. It should be noted that the components illustrated in the drawings are not necessarily drawn to scale. Descriptions of well-known components, processing techniques, and processes are omitted in this invention to avoid unnecessarily limiting the invention.

[0083] Example 1

[0084] like Figures 1-2 As shown, this invention provides a method for dynamic regulation of regional energy storage based on simulation models, including:

[0085] Step S1: Obtain historical data of equipment technical parameters, historical and real-time load data, historical data of energy storage system specifications, historical and real-time data of distributed photovoltaic power generation, and historical and real-time weather data for each transformer substation.

[0086] Step S3: Based on the historical data of equipment technical parameters, load historical data and real-time data, energy storage system specification historical data, distributed photovoltaic power generation historical data and real-time data, and weather historical data and real-time data in each transformer area, construct a simulation model.

[0087] Step S5: Based on the simulation model, monitor in real time whether each transformer area experiences reverse overload and whether the transformer area is currently in the peak period of distributed photovoltaic power generation. If the current transformer area meets the preset algorithm scheduling trigger conditions, according to the trigger condition type and the energy storage status of each transformer area, use the particle swarm optimization algorithm to generate the optimal charging strategy for the energy storage device. The preset algorithm scheduling trigger conditions include a first trigger condition and a second trigger condition. The first trigger condition is that a first preset number of transformer areas in the region all experience reverse overload, and the first preset number of transformer areas in the region are all in the peak period of distributed photovoltaic power generation. The second trigger condition is that the sum of the number of transformer areas in the region and the number of transformer areas in the peak period of distributed photovoltaic power generation is a second preset number, wherein the first preset number is less than the second preset number.

[0088] Step S7: Based on the generated optimal charging strategy for energy storage devices, perform dynamic energy storage scheduling for the energy storage devices in each distribution area to eliminate reverse heavy overload in the distribution area.

[0089] In step S1, data collection may include:

[0090] Power grid data: Collect data on all relevant equipment within the distribution area, including technical parameters of transformers, lines, switches, etc.

[0091] Load data: Collect historical load data and real-time load data for training and validation of load forecasting models;

[0092] Energy storage system data: Obtain the specifications of the energy storage system, including rated capacity, current stored power, charge and discharge rate, efficiency, etc.

[0093] Distributed photovoltaic data: Collects historical and real-time distributed photovoltaic power generation data;

[0094] Weather data: Collects weather forecast data.

[0095] In step S3, the simulation model first constructs a detailed power grid model, including a load model and a distributed photovoltaic (PV) power generation model, to simulate the power grid operation under various conditions. Next, through real-time power generation, the load data is calculated to determine whether the preset algorithm scheduling trigger conditions are met. When these conditions are met, the optimal charging strategy for energy storage devices is found using a particle swarm optimization algorithm. A comprehensive map of the regional power grid is created using simulation technology, including PV power generation devices, energy storage devices, power grid structure, load characteristics, etc., under each distribution area. Real-time power grid operation data, load data, and meteorological data are collected and displayed on this map.

[0096] Specifically, based on historical data of equipment technical parameters, historical and real-time load data, historical data of energy storage system specifications, historical and real-time data of distributed photovoltaic power generation, and historical and real-time weather data within each transformer substation area, a simulation model is constructed, including:

[0097] Based on historical data of equipment technical parameters and energy storage system specifications within each distribution area, a power grid model and an equipment electrical model are established. The power grid model includes a topology model of the power grid within different distribution areas, used to represent the connection relationships between various equipment nodes (such as transformers, lines, switches, etc.) within different distribution areas, showcasing the topology and electrical characteristics of the distribution network, providing an operating environment for other models, performing power flow calculations based on data predicted by the distributed photovoltaic power generation prediction model, and feeding the results back to the equipment model and load model. The equipment electrical model includes a distribution transformer model, a circuit breaker model, and a distributed power source model (such as photovoltaic panels), used to represent the electrical parameters of each device within different distribution areas, determining the role of the device in the power system based on the location information in the power grid model, displaying relevant electrical data of the device, and adjusting the status or output of the device based on the calculation results of the power grid model, such as adjusting the charging of energy storage equipment.

[0098] Based on historical and real-time load data, as well as historical and real-time weather data for each distribution area, a load model based on an LSTM neural network (reflecting users' electricity demand) is constructed. This model is used to predict the load curve of each distribution area (predicting the next few hours of the day) based on historical and real-time load data, historical weather data (such as light intensity and temperature), and real-time data. The load curve is then adjusted according to the calculation results of the power grid model to adapt to different operating conditions.

[0099] Based on historical and real-time data of distributed photovoltaic power generation in each transformer substation, as well as historical and real-time weather data, a distributed photovoltaic power generation prediction model based on LSTM neural network is constructed. This model is used to predict the distributed photovoltaic power generation of each transformer substation based on the historical and real-time data of distributed photovoltaic power generation and the historical and real-time weather data, and the prediction results are transmitted to the power grid model in the form of a file.

[0100] Simulation models can be used to evaluate grid performance under different conditions, helping to optimize operational strategies. For example, if a distributed photovoltaic (PV) power generation forecasting model for a certain area predicts high solar irradiance tomorrow, the PV system will generate more electricity. In this case, the grid model needs to recalculate the power flow distribution to ensure voltage stability and avoid overload. Simultaneously, the load model may predict higher electricity demand, and since more electricity is available, it may encourage the use of more power. Energy storage systems in the equipment model may be scheduled to store excess electricity.

[0101] The simulation model is used to simulate the operation of the entire regional power system, including simulating the working status of various electrical equipment (such as transformers, lines, switches, etc.) within each distribution area; predicting changes in voltage, current, power, and other factors under different operating conditions; analyzing the impact of the distribution area on the upstream power grid, and vice versa; predicting the load demand of the distribution area based on historical and real-time data, and analyzing the load characteristics at different time points (such as peak and off-peak periods); predicting the distributed photovoltaic power generation of the distribution area based on historical and real-time data, analyzing power generation at different time points, and combining this with load data for energy storage planning.

[0102] In this embodiment of the invention, the simulation model serves the following purposes: 1. To provide the initial environment configuration for the particle swarm optimization algorithm, displaying the voltage, current, and power status of each transformer area in real time. 2. To set parameters in the simulation model, such as the maximum charging and discharging power of the energy storage system and its energy storage capacity. 3. To define operational constraints for the energy storage system and other power system components, ensuring that the solution generated by the particle swarm optimization algorithm meets the constraints of actual operation. 4. To evaluate the effectiveness of the scheduling scheme proposed by the particle swarm optimization algorithm in the simulation environment and calculate the corresponding cost or benefit indicators as feedback information during the optimization process, guiding the particle swarm optimization algorithm to conduct further searches.

[0103] In step S5, overload refers to the phenomenon where the current or power carried by a component (conductor, transformer, etc.) in a power system exceeds its rated value. In some cases, loads in a power system can become power sources; for example, users with renewable energy generation systems may feed electricity back to the grid when they do not need all the generated power. This embodiment of the invention only considers renewable energy generation equipment present in the area, specifically photovoltaic power generation equipment. After the photovoltaic power generation meets the user's self-consumption needs, it will be fed back to the grid.

[0104] Based on simulation models, real-time monitoring is conducted to determine whether each transformer area experiences reverse overload and whether the area is currently in the peak period of distributed photovoltaic power generation. Specifically:

[0105] When current flows from the user to the power grid in a distribution area, and the transformer load rate is higher than a first preset value (e.g., 85%), then the distribution area experiences a reverse overload. The transformer load rate is calculated as follows:

[0106]

[0107] Where, β 反向 For transformer load factor; S 反向 This is the reverse apparent power (kVA) of the total meter for the distribution area; S e This is the rated capacity (kVA) of the transformer;

[0108] The simulation model displays the current, voltage, and power of each component and line in real time. When the current in the distribution area flows in the forward direction, the current flows from the upstream transformer through this node along the line to the load end. When the photovoltaic power generation in the distribution area is excessive, the current flows in the reverse direction to the grid. The simulation model shows the current flow in the reverse direction. The system automatically calculates according to the formula. When a reverse overload occurs, the reverse current is marked in red on the diagram, indicating a reverse overload.

[0109] If a distribution area is in a period of expected sufficient sunlight, and the rate of change of the power generation curve predicted by the corresponding distributed photovoltaic power generation prediction model for that area exceeds a preset rate of change threshold during the current period of expected sufficient sunlight, then the distribution area is currently in the peak period of distributed photovoltaic power generation.

[0110] This invention constructs a distributed photovoltaic (PV) power generation prediction model in a simulation model. This model is trained based on historical power generation data and meteorological data, and combines this with real-time weather conditions to predict PV power generation in subsequent time periods. Analysis of the predicted power generation curve reveals that during periods of expected abundant sunshine (e.g., between 10:00 AM and 2:00 PM), if the predicted PV power generation output by the distributed PV power generation prediction model is expected to show a significant increase (i.e., the rate of change of the predicted power generation curve exceeds a preset rate of change threshold), it may impact the power grid. In this situation, the system can issue commands to activate energy storage devices to store electrical energy, preventing reverse overload of the power grid and ensuring its normal operation.

[0111] In step S5, based on the trigger condition type and the energy storage status of each distribution area, the optimal charging strategy for the energy storage device is generated using the particle swarm optimization algorithm, specifically including:

[0112] Based on the triggering condition type and the energy storage situation of each distribution area, the objective function and constraints of the regulation are determined. The objective function is the sum of the grid overload and the charging and discharging cost of the energy storage device. The constraints are that each distribution area does not experience reverse heavy overload or is not currently in the peak period of distributed photovoltaic power generation.

[0113] A particle swarm optimization algorithm with added constraints is introduced to generate the optimal charging strategy for energy storage devices.

[0114] Wherein, when the triggering condition type is the first triggering condition (for example, 30% of the transformer substations in the region experience reverse heavy overload, and 30% of the transformer substations in the region are in the peak period of distributed photovoltaic power generation), the objective function is:

[0115]

[0116] Where Objective 1 is the minimum sum of grid overload and energy storage device charging / discharging costs under the first triggering condition; N1 is the number of distribution substations experiencing reverse overload and during peak distributed photovoltaic (PV) generation; Q1 is the number of distribution substations not experiencing reverse overload and not during peak distributed PV generation; M is the number of energy storage devices within the distribution substation; PV excess,i It refers to the excess photovoltaic power generation of transformer area i, which is experiencing reverse overload and is in the peak period of distributed photovoltaic power generation. Cost is the amount of electricity allocated to energy storage devices j within transformer area α1 that have not experienced reverse overload and are not in the peak period of distributed photovoltaic power generation; Allocate,αj It is the total cost allocated to the energy storage device j in the distribution area α1 that has not experienced reverse heavy overload and is not in the peak period of distributed photovoltaic power generation. The total cost includes charging and discharging loss cost and equipment depreciation cost.

[0117] When the triggering condition type is the second triggering condition (e.g., 60% of the transformer substations in the region experience reverse heavy overload, or 60% of the transformer substations in the region are in the peak period of distributed photovoltaic power generation), the objective function is:

[0118]

[0119] Where Objective2 is the minimum sum of grid overload and energy storage device charging / discharging costs when the trigger condition is met; N2 is the number of distribution substations experiencing reverse overload or during peak distributed photovoltaic (PV) generation; Q2 is the number of distribution substations without reverse overload (Q2 is greater than Q1, meaning that distribution substations during peak distributed PV generation can be included in the allocation, as long as reverse overload does not occur); M is the number of energy storage devices within the distribution substation; PV excess,i It refers to the excess photovoltaic power generation of transformer area i, which is experiencing reverse overload and is in the peak period of distributed photovoltaic power generation. It is the amount of electricity allocated to the energy storage device j in the transformer area α2 where no reverse overload has occurred; It is the total cost allocated to energy storage device j in transformer area α2 where no reverse overload has occurred. The total cost includes charging and discharging loss costs and equipment depreciation costs.

[0120] When the triggering condition is either the first or the second triggering condition, it will affect the entire distribution area or the area upstream of the distribution area. Based on the energy storage situation of each distribution area, the particle swarm optimization algorithm will be used for dynamic energy storage scheduling to ensure the normal operation of the entire regional power grid.

[0121] Furthermore, when the triggering condition type is the first triggering condition, the constraints all include grid carrying capacity constraints and energy storage device capacity constraints, wherein the grid carrying capacity constraints are:

[0122]

[0123] Among them, L t TotalLoad represents the grid carrying capacity at time t. t This represents the total load of the power grid at time t;

[0124] The capacity constraints for energy storage devices are:

[0125]

[0126] Among them, C max,j It is the maximum capacity that the j-th energy storage device can store;

[0127] When the trigger condition type is the second trigger condition, the constraints all include grid carrying capacity constraints and energy storage device capacity constraints. Among them, the grid carrying capacity constraints are:

[0128]

[0129] The capacity constraints for energy storage devices are:

[0130]

[0131] Preferably, if a constraint is violated, the objective function is adjusted using a penalty function;

[0132] The penalty function includes the equipment capacity penalty function Penalty. capacity,αj Penalty function for grid carrying capacity load,t When the trigger condition type is the first trigger condition, the device capacity penalty function The expression is:

[0133]

[0134] When the trigger condition type is the second trigger condition, the device capacity penalty function The expression is:

[0135]

[0136] When the trigger condition type is the first trigger condition, the grid carrying capacity penalty function Penalty load,t,1 The expression is:

[0137]

[0138] When the triggering condition type is the second triggering condition, the grid carrying capacity penalty function Penalty load,t,2 The expression is:

[0139]

[0140] When the trigger condition type is the first trigger condition, the adjusted new objective function is:

[0141]

[0142] Where λ is the penalty function coefficient, used to control the influence of the penalty function on the objective function;

[0143] When the trigger condition type is the second trigger condition, the adjusted new objective function is:

[0144]

[0145] Where λ is the penalty function coefficient, used to control the influence of the penalty function on the objective function.

[0146] Specifically, the introduction of a particle swarm optimization algorithm with added constraints to generate the optimal charging strategy for energy storage devices includes:

[0147] Initialize the particle position parameters. The position of each particle can be represented as an M×N matrix, where each element... This represents the proportion or absolute value of the electricity allocated to energy storage device j in distribution area α1, or, for each element. This represents the proportion or absolute value of the electricity allocated to energy storage device j in transformer area α2;

[0148] Calculate the fitness value, where the fitness value is the objective function value;

[0149] Determine if any particle violates the constraints. If a particle violates the constraints, apply the penalty function to adjust the objective function, and recalculate the fitness value of each particle based on the adjusted objective function.

[0150] Based on the individual extreme value and global extreme value of each particle, update the particle position and velocity, where the update expression for particle position and velocity is:

[0151] v mn (t+1)=a(t)v mn (t)+b1r1(t)(p mn (t)-y mn (t))+b2r2(t)(p gn (t)-y mn (t)),

[0152] y mn (t+1)=y mn (t)+v mn (t+1),

[0153] Among them, v mn (t), y mn (t), p mn (t) represents the velocity, position, and individual extreme value of the m-th particle in dimension n at the t-th iteration; p gn (t) represents the global extremum in dimension n at the t-th iteration; a(t) represents the magnitude of the inertia weight at the t-th iteration; b1 and b2 are learning factors; r1(t) and r2(t) are random numbers in the interval (0,1].

[0154] The position and velocity of each particle are iteratively updated until a preset stopping condition is met. The preset stopping condition includes reaching the maximum number of iterations or the fitness value being greater than a preset fitness threshold.

[0155] The particle swarm optimization algorithm is used to optimize the charging and discharging strategy of the energy storage system, so as to maximize the effective allocation of excess photovoltaic power generation to the energy storage device and eliminate grid overload as a comprehensive objective. The objective function is to minimize grid overload and the charging and discharging cost of the energy storage device. The optimal solution is obtained by introducing the particle swarm optimization algorithm with added constraints.

[0156] The allocated power is primarily based on the storable energy of each energy storage device and the device's state (charging, discharging). Devices are not considered when they are discharging. Storable energy is used as a constraint.

[0157] Among them, a(t)=a(t) max -t(a(t) max -a(t) min ) / t max a(t) represents the magnitude of the inertia weight in the t-th iteration, which can be adjusted. max The maximum number of iterations,

[0158] like Figure 3 As shown, the technical solution of the present invention also provides a regional energy storage dynamic control system based on simulation model. Before performing dynamic energy storage scheduling on the energy storage devices in each substation area according to the generated optimal charging strategy of the energy storage devices in step S7, the system further includes:

[0159] S6. Simulate and verify the generated optimal charging strategy for the energy storage device in the simulation model. If the reverse overload phenomenon disappears in each area, the optimal charging strategy for the energy storage device is verified and step S7 is executed. If the reverse overload phenomenon does not disappear in a certain area, the optimal charging strategy for the energy storage device is not verified and step S5 is executed.

[0160] The suitability of the allocation method can be verified by using simulation models. That is, the scheduling scheme (optimal charging strategy for energy storage devices) is simulated in the simulation model. If the reverse overload phenomenon in each area disappears, the scheduling scheme can be implemented.

[0161] This invention constructs a simulation model based on historical data of equipment technical parameters, load history and real-time data, energy storage system specifications, distributed photovoltaic power generation history and real-time data, and weather history and real-time data for each power distribution area. Based on this simulation model, it monitors in real-time whether each power distribution area experiences reverse overload and whether it is currently in a peak period of distributed photovoltaic power generation. According to the triggering condition type and the energy storage status of each power distribution area, it uses a particle swarm optimization algorithm to generate the optimal charging strategy for the energy storage devices. Based on the generated optimal charging strategy, it performs dynamic energy storage scheduling for the energy storage devices in each power distribution area to eliminate reverse overload. By combining the accurate simulation of the simulation model with the efficient search capability of the particle swarm optimization algorithm, it achieves precise control and optimized scheduling of regional power grid energy storage devices, effectively solving the problems of low reliability and economy in energy storage regulation caused by existing technologies, effectively saving regulation costs and improving the reliability of energy storage regulation.

[0162] The simulation model in the technical solution of this invention includes a power grid model, an equipment electrical model, a load model, and a distributed photovoltaic power generation prediction model. These models are interconnected and work together to respond to changes in power grid load in real time, intelligently allocate and adjust energy storage resources, effectively alleviate the peak-valley difference problem of the power grid, and improve the operating efficiency and economy of the power grid.

[0163] In this invention's technical solution, real-time monitoring based on simulation models indicates that when current flows from the user to the grid in a distribution area, and the transformer load rate is higher than a first preset value, the distribution area experiences reverse overload. This not only considers the grid model in the simulation model but also incorporates the electrical parameters in the equipment's electrical model, thus organically combining efficiency and accuracy. When a distribution area is in a period of expected sufficient sunlight, and the rate of change of the power generation curve predicted by the corresponding distributed photovoltaic power generation prediction model for that area exceeds a preset rate of change threshold during the current period of expected sufficient sunlight, then the distribution area is currently in the peak period of distributed photovoltaic power generation. This not only considers the time period of the distribution area in the simulation model but also incorporates the output prediction results of the distributed photovoltaic power generation prediction model, thus organically combining efficiency and accuracy.

[0164] In this invention, the objective function and constraints for regulation are determined based on the triggering condition type and the energy storage status of each distribution area. The objective function is the sum of the grid overload and the charging and discharging costs of the energy storage device, and the constraint is that no reverse heavy overload occurs in any distribution area. A particle swarm optimization algorithm with added constraints is introduced to generate the optimal charging strategy for the energy storage device. If the constraint is violated, the objective function is adjusted through a penalty function. Not only does the objective function vary depending on the triggering condition type, but the penalty function is also adjusted when the constraint is violated, improving the adaptability of energy storage regulation. Furthermore, by combining the accurate simulation of the simulation model with the efficient search capability of the particle swarm optimization algorithm, precise control and optimized scheduling of regional power grid energy storage devices are achieved. This enhances the adaptability to renewable energy fluctuations and the grid's anti-interference capability, reduces energy waste caused by supply and demand imbalances, and provides an innovative solution for the intelligent management and sustainable development of the power grid.

[0165] Before dynamically scheduling energy storage for each distribution area based on the generated optimal charging strategy, the technical solution of this invention further includes: simulating and verifying the generated optimal charging strategy in a simulation model. If the reverse overload phenomenon disappears in each distribution area, the optimal charging strategy is verified successfully; if the reverse overload phenomenon does not disappear in a certain distribution area, the optimal charging strategy is not verified successfully. This ensures the effectiveness of the implementation of the optimal charging strategy and improves the economic efficiency of energy storage regulation.

[0166] Example 2

[0167] like Figure 4 As shown, the present invention also provides a regional energy storage dynamic control system based on simulation model, comprising:

[0168] Module 101 acquires historical data of equipment technical parameters, load, energy storage system specifications, distributed photovoltaic power generation, and weather for each transformer substation.

[0169] Module 102 constructs a simulation model based on historical data of equipment technical parameters, historical and real-time data of load, historical data of energy storage system specifications, historical and real-time data of distributed photovoltaic power generation, and historical and real-time data of weather in each transformer area.

[0170] The monitoring and generation module 103, based on the simulation model, monitors in real time whether each transformer area experiences reverse overload and whether the transformer area is currently in the peak period of distributed photovoltaic power generation. If the current transformer area meets the preset algorithm scheduling trigger conditions, it generates the optimal charging strategy for the energy storage device using the particle swarm optimization algorithm according to the trigger condition type and the energy storage status of each transformer area. The preset algorithm scheduling trigger conditions include a first trigger condition and a second trigger condition. The first trigger condition is that a first preset number of transformer areas in the region all experience reverse overload, and the first preset number of transformer areas in the region are all in the peak period of distributed photovoltaic power generation. The second trigger condition is that the sum of the number of transformer areas in the region and the number of transformer areas in the peak period of distributed photovoltaic power generation is a second preset number, wherein the first preset number is less than the second preset number.

[0171] The scheduling module 104 performs dynamic energy storage scheduling for the energy storage devices in each distribution area based on the generated optimal charging strategy for the energy storage devices, in order to eliminate reverse heavy overload in the distribution area.

[0172] The acquisition module 101, construction module 102, monitoring and generation module 103, and scheduling module 104 in the technical solution of this invention correspond to the steps in the embodiment, and will not be described in detail here.

[0173] This invention constructs a simulation model based on historical data of equipment technical parameters, load history and real-time data, energy storage system specifications, distributed photovoltaic power generation history and real-time data, and weather history and real-time data for each power distribution area. Based on this simulation model, it monitors in real-time whether each power distribution area experiences reverse overload and whether it is currently in a peak period of distributed photovoltaic power generation. According to the triggering condition type and the energy storage status of each power distribution area, it uses a particle swarm optimization algorithm to generate the optimal charging strategy for the energy storage devices. Based on the generated optimal charging strategy, it performs dynamic energy storage scheduling for the energy storage devices in each power distribution area to eliminate reverse overload. By combining the accurate simulation of the simulation model with the efficient search capability of the particle swarm optimization algorithm, it achieves precise control and optimized scheduling of regional power grid energy storage devices, effectively solving the problems of low reliability and economy in energy storage regulation caused by existing technologies, effectively saving regulation costs and improving the reliability of energy storage regulation.

[0174] The simulation model in the technical solution of this invention includes a power grid model, an equipment electrical model, a load model, and a distributed photovoltaic power generation prediction model. These models are interconnected and work together to respond to changes in power grid load in real time, intelligently allocate and adjust energy storage resources, effectively alleviate the peak-valley difference problem of the power grid, and improve the operating efficiency and economy of the power grid.

[0175] In this invention's technical solution, real-time monitoring based on simulation models indicates that when current flows from the user to the grid in a distribution area, and the transformer load rate is higher than a first preset value, the distribution area experiences reverse overload. This not only considers the grid model in the simulation model but also incorporates the electrical parameters in the equipment's electrical model, thus organically combining efficiency and accuracy. When a distribution area is in a period of expected sufficient sunlight, and the rate of change of the power generation curve predicted by the corresponding distributed photovoltaic power generation prediction model for that area exceeds a preset rate of change threshold during the current period of expected sufficient sunlight, then the distribution area is currently in the peak period of distributed photovoltaic power generation. This not only considers the time period of the distribution area in the simulation model but also incorporates the output prediction results of the distributed photovoltaic power generation prediction model, thus organically combining efficiency and accuracy.

[0176] In this invention, the objective function and constraints for regulation are determined based on the triggering condition type and the energy storage status of each distribution area. The objective function is the sum of the grid overload and the charging and discharging costs of the energy storage device, and the constraint is that no reverse heavy overload occurs in any distribution area. A particle swarm optimization algorithm with added constraints is introduced to generate the optimal charging strategy for the energy storage device. If the constraint is violated, the objective function is adjusted through a penalty function. Not only does the objective function vary depending on the triggering condition type, but the penalty function is also adjusted when the constraint is violated, improving the adaptability of energy storage regulation. Furthermore, by combining the accurate simulation of the simulation model with the efficient search capability of the particle swarm optimization algorithm, precise control and optimized scheduling of regional power grid energy storage devices are achieved. This enhances the adaptability to renewable energy fluctuations and the grid's anti-interference capability, reduces energy waste caused by supply and demand imbalances, and provides an innovative solution for the intelligent management and sustainable development of the power grid.

[0177] Before dynamically scheduling energy storage for each distribution area based on the generated optimal charging strategy, the technical solution of this invention further includes: simulating and verifying the generated optimal charging strategy in a simulation model. If the reverse overload phenomenon disappears in each distribution area, the optimal charging strategy is verified successfully; if the reverse overload phenomenon does not disappear in a certain distribution area, the optimal charging strategy is not verified successfully. This ensures the effectiveness of the implementation of the optimal charging strategy and improves the economic efficiency of energy storage regulation.

[0178] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for dynamic regulation of regional energy storage based on simulation models, characterized in that, include: Acquire historical data on equipment technical parameters, load, and real-time data for each transformer substation, as well as historical and real-time data on energy storage system specifications, distributed photovoltaic power generation, and weather. Based on historical data of equipment technical parameters, load history and real-time data, energy storage system specifications history data, distributed photovoltaic power generation history and real-time data, and weather history and real-time data for each transformer substation, a simulation model is constructed; specifically including: Based on historical data of equipment technical parameters and energy storage system specifications within each distribution area, a power grid model and an equipment electrical model are established. The power grid model includes a topology model of the power grid within different distribution areas, used to represent the connection relationships between various equipment nodes within different distribution areas, displaying the topology and electrical characteristics of the distribution network, providing an operating environment for other models, performing power flow calculations based on data predicted by the distributed photovoltaic power generation prediction model, and feeding the results back to the equipment model and load model. The equipment electrical model includes a distribution transformer model, a circuit breaker model, and a distributed power source model, used to represent the electrical parameters of each device within different distribution areas, determining the role of the device in the power system based on the location information in the power grid model, displaying relevant electrical data of the device, and adjusting the status or output of the device based on the calculation results of the power grid model. Based on historical and real-time load data, as well as historical and real-time weather data for each distribution area, a load model based on an LSTM neural network is constructed. This model is used to predict the load curves of each distribution area based on the historical and real-time load data, as well as the historical and real-time weather data. The load curves are then adjusted according to the calculation results of the power grid model to adapt to different operating conditions. Based on historical and real-time data of distributed photovoltaic power generation in each transformer area, as well as historical and real-time weather data, a distributed photovoltaic power generation prediction model based on LSTM neural network is constructed. This model is used to predict the distributed photovoltaic power generation of each transformer area based on historical and real-time data of distributed photovoltaic power generation, as well as historical and real-time weather data, and the prediction results are transmitted to the power grid model in the form of a file. Based on the simulation model, the system monitors in real time whether each transformer area experiences reverse overload and whether it is currently in the peak period of distributed photovoltaic power generation. If the current transformer area meets the preset algorithm scheduling trigger conditions, the optimal charging strategy for the energy storage device is generated using the particle swarm optimization algorithm according to the trigger condition type and the energy storage status of each transformer area. The preset algorithm scheduling trigger conditions include a first trigger condition and a second trigger condition. The first trigger condition is that a first preset number of transformer areas in the region all experience reverse overload, and the first preset number of transformer areas in the region are all in the peak period of distributed photovoltaic power generation. The second trigger condition is that the sum of the number of transformer areas experiencing reverse overload and the number of transformer areas in the peak period of distributed photovoltaic power generation is a second preset number, wherein the first preset number is less than the second preset number. Based on the generated optimal charging strategy for energy storage devices, dynamic energy storage scheduling is performed on the energy storage devices in each distribution area to eliminate reverse overload in the distribution area.

2. The method for dynamic regulation of regional energy storage based on simulation models according to claim 1, characterized in that, Based on simulation models, real-time monitoring is conducted to determine whether each transformer area experiences reverse overload and whether the area is currently in the peak period of distributed photovoltaic power generation. Specifically: When current flows from the user to the power grid in a distribution area, and the transformer load rate is higher than a first preset value, then the distribution area experiences a reverse overload; the transformer load rate is calculated as follows: Where, β 反向 For transformer load factor; S 反向 This is the reverse apparent power of the total meter for the distribution area; S e This is the rated capacity of the transformer; When a distribution area is in a period of expected sufficient sunlight, and the rate of change of the power generation curve predicted by the corresponding distributed photovoltaic power generation prediction model for that area exceeds a preset rate of change threshold during the current period of expected sufficient sunlight, then the distribution area is currently in the peak period of distributed photovoltaic power generation.

3. The method for dynamic regulation of regional energy storage based on simulation model according to claim 1, characterized in that, Based on the trigger condition type and the energy storage status of each transformer area, the optimal charging strategy for the energy storage device is generated using the particle swarm optimization algorithm, specifically including: Based on the triggering condition type and the energy storage situation of each distribution area, the objective function and constraints of regulation are determined. The objective function is the sum of the grid overload and the charging and discharging cost of the energy storage device, and the constraint is that no reverse heavy overload occurs in each distribution area. A particle swarm optimization algorithm with added constraints is introduced to generate the optimal charging strategy for energy storage devices.

4. The method for dynamic regulation of regional energy storage based on simulation model according to claim 3, characterized in that, When the trigger condition type is the first trigger condition, the objective function is: Where Objective 1 is the minimum sum of grid overload and energy storage device charging / discharging costs under the first triggering condition; N1 is the number of distribution substations experiencing reverse overload and during peak distributed photovoltaic (PV) generation; Q1 is the number of distribution substations not experiencing reverse overload and not during peak distributed PV generation; M is the number of energy storage devices within the distribution substation; PV excess,i It refers to the excess photovoltaic power generation of transformer area i, which is experiencing reverse overload and is in the peak period of distributed photovoltaic power generation. It is the amount of electricity allocated to the energy storage device j in the distribution area α1 that has not experienced reverse overload and is not in the peak period of distributed photovoltaic power generation; It is the total cost allocated to the energy storage device j in the distribution area α1 that has not experienced reverse heavy overload and is not in the peak period of distributed photovoltaic power generation. The total cost includes charging and discharging loss cost and equipment depreciation cost. When the trigger condition type is the second trigger condition, the objective function is: Where Objective2 is the minimum sum of grid overload and energy storage device charging / discharging costs under the second triggering condition; N2 is the number of distribution substations experiencing reverse overload or during peak distributed photovoltaic power generation; Q2 is the number of distribution substations without reverse overload; M is the number of energy storage devices within the distribution substation; PV excess,i It refers to the excess photovoltaic power generation of transformer area i, which is experiencing reverse overload and is in the peak period of distributed photovoltaic power generation. It is the amount of electricity allocated to the energy storage device j in the transformer area α2 where no reverse overload has occurred; It is the total cost allocated to energy storage device j in transformer area α2 where no reverse overload has occurred. The total cost includes charging and discharging loss costs and equipment depreciation costs.

5. The method for dynamic regulation of regional energy storage based on simulation model according to claim 4, characterized in that, When the trigger condition type is the first trigger condition, the constraints all include grid carrying capacity constraints and energy storage device capacity constraints. Among them, the grid carrying capacity constraints are: Among them, L t TotalLoad represents the grid carrying capacity at time t. t This represents the total load of the power grid at time t; The capacity constraints for energy storage devices are: Among them, C max,j It is the maximum capacity that the j-th energy storage device can store; When the trigger condition type is the second trigger condition, the constraints all include grid carrying capacity constraints and energy storage device capacity constraints. Among them, the grid carrying capacity constraints are: The capacity constraints for energy storage devices are:

6. The method for dynamic regulation of regional energy storage based on simulation model according to claim 5, characterized in that, If a constraint is violated, the objective function is adjusted using a penalty function. The penalty function includes the equipment capacity penalty function Penalty. capacity,αj Penalty function for grid carrying capacity load,t When the trigger condition type is the first trigger condition, the device capacity penalty function The expression is: When the trigger condition type is the second trigger condition, the device capacity penalty function The expression is: When the trigger condition type is the first trigger condition, the grid carrying capacity penalty function Penalty load,t,1 The expression is: When the triggering condition type is the second triggering condition, the grid carrying capacity penalty function Penalty load,t,2 The expression is: When the trigger condition type is the first trigger condition, the adjusted new objective function is: Where λ is the penalty function coefficient, used to control the influence of the penalty function on the objective function; When the trigger condition type is the second trigger condition, the adjusted new objective function is:

7. The method for dynamic regulation of regional energy storage based on simulation model according to claim 3, characterized in that, it introduces... The particle swarm optimization algorithm with added constraints, which generates the optimal charging strategy for energy storage devices, specifically includes: Initialize the particle position parameters. The position of each particle can be represented as an M×N matrix, where each element... This represents the proportion or absolute value of the electricity allocated to energy storage device j in distribution area α1, or, for each element. This represents the proportion or absolute value of the electricity allocated to energy storage device j in transformer area α2; Calculate the fitness value, where the fitness value is the objective function value; Determine if any particle violates the constraints. If a particle violates the constraints, apply the penalty function to adjust the objective function, and recalculate the fitness value of each particle based on the adjusted objective function. Based on the individual extreme value and global extreme value of each particle, update the particle position and velocity, where the update expression for particle position and velocity is: v mn (t+1)=a(t)v mn (t)+b1r1(t)(p mn (t)-y mn (t))+b2r2(t)(p gn (t)-y mn (t)), y mn (t+1)=y mn (t)+v mn (t+1), Among them, v mn (t), y mn (t), p mn (t) represents the velocity, position, and individual extreme value of the m-th particle in dimension n at the t-th iteration; p gn (t) represents the global extremum in dimension n at the t-th iteration; a(t) represents the magnitude of the inertia weight at the t-th iteration; b1 and b2 are learning factors; r1(t) and r2(t) are random numbers in the interval (0,1]. The position and velocity of each particle are iteratively updated until a preset stopping condition is met. The preset stopping condition includes reaching the maximum number of iterations or the fitness value being greater than a preset fitness threshold.

8. A method for dynamic regulation of regional energy storage based on simulation models according to any one of claims 1-7, characterized in that, Before performing dynamic energy storage scheduling on the energy storage devices in each distribution area based on the generated optimal charging strategy for the energy storage devices, the following steps are also included: The generated optimal charging strategy for the energy storage device is simulated and verified in the simulation model. If the reverse overload phenomenon disappears in all areas, the optimal charging strategy for the energy storage device is verified. If the reverse overload phenomenon does not disappear in a certain area, the optimal charging strategy for the energy storage device is not verified.

9. A regional energy storage dynamic control system based on simulation model, characterized in that, include: The acquisition module acquires historical data on equipment technical parameters, load, energy storage system specifications, distributed photovoltaic power generation, and weather for each transformer substation. The construction module, based on historical data of equipment technical parameters, historical and real-time load data, historical data of energy storage system specifications, historical and real-time data of distributed photovoltaic power generation, and historical and real-time weather data within each transformer substation, constructs simulation models; specifically including: Based on historical data of equipment technical parameters and energy storage system specifications within each distribution area, a power grid model and an equipment electrical model are established. The power grid model includes a topology model of the power grid within different distribution areas, used to represent the connection relationships between various equipment nodes within different distribution areas, displaying the topology and electrical characteristics of the distribution network, providing an operating environment for other models, performing power flow calculations based on data predicted by the distributed photovoltaic power generation prediction model, and feeding the results back to the equipment model and load model. The equipment electrical model includes a distribution transformer model, a circuit breaker model, and a distributed power source model, used to represent the electrical parameters of each device within different distribution areas, determining the role of the device in the power system based on the location information in the power grid model, displaying relevant electrical data of the device, and adjusting the status or output of the device based on the calculation results of the power grid model. Based on historical and real-time load data, as well as historical and real-time weather data for each distribution area, a load model based on an LSTM neural network is constructed. This model is used to predict the load curves of each distribution area based on the historical and real-time load data, as well as the historical and real-time weather data. The load curves are then adjusted according to the calculation results of the power grid model to adapt to different operating conditions. Based on historical and real-time data of distributed photovoltaic power generation in each transformer area, as well as historical and real-time weather data, a distributed photovoltaic power generation prediction model based on LSTM neural network is constructed. This model is used to predict the distributed photovoltaic power generation of each transformer area based on historical and real-time data of distributed photovoltaic power generation, as well as historical and real-time weather data, and the prediction results are transmitted to the power grid model in the form of a file. The monitoring and generation module, based on simulation models, monitors in real time whether each transformer area experiences reverse overload and whether the area is currently in the peak period of distributed photovoltaic power generation. If the current transformer area meets the preset algorithm scheduling trigger conditions, it generates the optimal charging strategy for the energy storage device using the particle swarm optimization algorithm, based on the trigger condition type and the energy storage status of each transformer area. The preset algorithm scheduling trigger conditions include a first trigger condition and a second trigger condition. The first trigger condition is that a first preset number of transformer areas in the region all experience reverse overload, and the first preset number of transformer areas in the region are all in the peak period of distributed photovoltaic power generation. The second trigger condition is that the sum of the number of transformer areas in the region and the number of transformer areas in the peak period of distributed photovoltaic power generation is a second preset number, wherein the first preset number is less than the second preset number. The scheduling module performs dynamic energy storage scheduling for each distribution area based on the generated optimal charging strategy for the energy storage devices, in order to eliminate reverse overload in the distribution area.

Images