Energy storage system control method and system based on time-of-use electricity price and multi-modal optimization

Abstract

The invention discloses an energy storage system control method and system based on time-of-use electricity price and multi-modal optimization, and the method comprises the steps: simulating supply and demand fluctuation in a complex scene through a pre-constructed energy storage system control strategy model based on deep reinforcement learning, dynamically optimizing a charging and discharging plan of energy storage equipment, and generating an energy storage resource distribution efficiency optimization scheme; generating a real-time electricity price adjustment scheme by adopting a game theory price adjustment algorithm based on the energy storage resource distribution efficiency optimization scheme, the real-time demand of power grid supply and demand balance, the user electricity consumption behavior change and the power grid operation state dynamic state; generating a user behavior response data set through a pre-constructed user behavior response model based on the real-time electricity price adjustment scheme in combination with user economic benefit protection requirements; and calculating a fluctuation ratio of a power grid load change trend by combining the user behavior response data set with real-time monitoring data through a power grid supply and demand balance dynamic evaluation algorithm, and determining a final energy storage scheduling scheme. The purposes of real-time monitoring and collaborative optimization of energy storage system control are achieved.

Description

Technical Field

[0001] This invention relates to the field of information technology, and specifically to a control method and system for energy storage systems based on time-of-use pricing and multimodal optimization. Background Technology

[0002] In the fields of energy management and smart grids, the optimized control of energy storage systems and the rational design of electricity pricing mechanisms are key pillars for promoting efficient energy utilization and stable grid operation. Research in this area is directly related to the success or failure of energy transition and has irreplaceable significance for ensuring the security and economy of the power system. With the widespread integration of renewable energy and the diversification of user demands, the role of energy storage systems in balancing grid supply and demand and improving energy utilization efficiency is becoming increasingly prominent.

[0003] However, current solutions generally have limitations, primarily in their insufficient comprehensive consideration of the power grid, energy storage devices, and user behavior. Many methods tend to focus only on optimizing a single aspect, lacking the overall coordination capability for dynamic changes across multiple parties. This makes them difficult to adapt to complex operating environments in practical applications, thereby affecting the stability of the power grid and the economic interests of users.

[0004] A deeper analysis of the challenges in this field reveals that the core issues lie in the real-time integration and collaborative optimization of multi-dimensional information. Firstly, due to the highly dynamic and heterogeneous nature of grid status, energy storage device status, and user behavior data, existing technologies struggle to achieve comprehensive real-time monitoring and fusion of this information, resulting in incomplete decision-making basis. This lack of information integration further leads to insufficient adaptability of optimization algorithms in complex scenarios, failing to effectively balance the contradiction between grid demand and energy storage resource allocation. At a deeper level, the lack of a real-time response mechanism to grid regulation demands means that the dynamic generation of electricity prices often lags behind actual demand, making it difficult to truly guide user behavior and ensure grid security.

[0005] Therefore, how to construct a dynamic electricity pricing mechanism that can monitor and coordinately optimize the control of energy storage systems in real time and generate dynamic electricity pricing that adapts to the needs of the power grid in an environment of dynamic changes in multi-dimensional information has become a key problem that this research needs to solve. Summary of the Invention

[0006] To address the challenge of effectively balancing grid demand and energy storage resource allocation in complex scenarios, and to mitigate the problem that dynamic electricity pricing often lags behind actual demand due to the lack of a real-time response mechanism for grid regulation, thus failing to truly guide user behavior and ensure grid security, this invention proposes an energy storage system control method based on time-of-use pricing and multimodal optimization, comprising:

[0007] Based on a pre-built energy storage system control strategy model, the system simulates supply and demand fluctuations under complex scenarios, dynamically optimizes the charging and discharging plans of energy storage devices, and generates an energy storage resource allocation efficiency optimization scheme.

[0008] Based on the energy storage resource allocation efficiency optimization scheme, the real-time demand for power grid supply and demand balance, changes in user electricity consumption behavior and the dynamics of power grid operation status, a game theory price adjustment algorithm is used to generate a real-time electricity price adjustment scheme.

[0009] Based on the real-time electricity price adjustment scheme and the need to protect users' economic interests, a user behavior response dataset is generated through a pre-built user behavior response model.

[0010] Based on user behavior response datasets and real-time monitoring data of power grid operation status, the volatility of power grid load change trend is calculated using a power grid supply and demand balance dynamic evaluation algorithm; and based on whether the volatility of the power grid load change trend reaches a preset stability threshold, the final energy storage dispatch scheme is determined.

[0011] Optionally, the construction of the energy storage system control strategy model includes:

[0012] Data on power grid operation status, energy storage device operation, and user electricity consumption behavior are acquired, and the data on power grid operation status, energy storage device operation, and user electricity consumption behavior are cleaned, standardized, and formatted to obtain a power grid operation dataset.

[0013] Data fusion technology is used to fuse the data in the power grid operation dataset, and a deep reinforcement learning algorithm is used to extract the time-series correlation features of multi-dimensional information to generate a power grid operation correlation feature matrix.

[0014] By utilizing the long short-term memory network's ability to capture long-term dependencies on time-series data, the correlation feature matrix of power grid operation is used to predict the future trends of power grid operation status and user electricity consumption behavior, generate the power grid supply and demand balance demand value and the available capacity range of energy storage devices within the prediction period, and form the power grid operation prediction demand distribution.

[0015] Based on the distribution of demand predicted by power grid operation and the available capacity range of energy storage devices, a control strategy model for energy storage systems based on deep reinforcement learning is constructed.

[0016] Optionally, the method of simulating supply and demand fluctuations under complex scenarios based on a pre-built energy storage system control strategy model, dynamically optimizing the charging and discharging plans of energy storage devices, and generating an energy storage resource allocation efficiency optimization scheme includes:

[0017] The pre-built energy storage system control strategy model is used to simulate the operation state under complex scenarios and determine the operation constraints of energy storage devices in different time periods.

[0018] Based on the analysis of simulation results, a control strategy framework is constructed to meet the needs of dynamic adjustment.

[0019] To obtain the optimal response mode of energy storage equipment under supply and demand fluctuations, and to obtain a preliminary scheme for charging and discharging;

[0020] Based on whether the preliminary scheme exceeds the available capacity limit of the energy storage device, determine whether to adjust the parameters of the control strategy framework to obtain an updated scheme that meets the available capacity limit;

[0021] Based on the updated scheme and real-time data of power grid operation, short-term changes in supply and demand fluctuations are monitored. If abnormal fluctuations are detected, the parameters of the control strategy framework are adjusted to optimize the updated scheme and obtain an optimized scheme for energy storage resource allocation efficiency. Otherwise, the updated scheme is directly used as the optimized scheme for energy storage resource allocation efficiency.

[0022] Optionally, the real-time electricity price adjustment scheme is generated based on the energy storage resource allocation efficiency optimization scheme, the real-time demand of power grid supply and demand balance, changes in user electricity consumption behavior, and the dynamics of power grid operation status, using a game theory price adjustment algorithm, including:

[0023] Based on the energy storage resource allocation efficiency optimization scheme and the real-time demand of power grid supply and demand balance, a dynamic electricity price generation model is constructed. A game theory price adjustment algorithm based on Nash equilibrium is adopted to generate a dynamic electricity price adjustment scheme by combining user electricity consumption behavior, power grid operation dynamics, and power grid supply and demand balance status.

[0024] Based on the price change trend of the aforementioned dynamic electricity price adjustment scheme, the energy storage resource allocation efficiency optimization scheme is adjusted, and the resource allocation and electricity price linkage scheme is determined.

[0025] The parameters in the dynamic electricity price generation model are updated by the resource allocation and electricity price linkage scheme. Based on the updated dynamic electricity price generation model, the real-time demand for power grid supply and demand balance is monitored, and a real-time electricity price adjustment scheme is generated.

[0026] Optionally, the dynamic electricity price generation model is constructed based on the energy storage resource allocation efficiency optimization scheme and the real-time demand of power grid supply and demand balance. A game-theoretic price adjustment algorithm based on Nash equilibrium is used to generate a dynamic electricity price adjustment scheme by combining user electricity consumption behavior, power grid operation dynamics, and power grid supply and demand balance status. This scheme includes:

[0027] Acquire power grid supply and demand data and balance status information, collect electricity load data and energy storage resource status data in the current period, perform data cleaning and formatting on the electricity load data and energy storage resource status data to obtain a standardized supply and demand dataset;

[0028] The matching degree between the energy storage resource allocation efficiency and the power grid supply and demand data is analyzed based on the supply and demand dataset. When the matching degree is lower than the preset threshold, the allocation efficiency optimization process is triggered. By comparing the operation dynamics in historical data, the adjustment direction of the current energy storage resources is determined.

[0029] Based on the adjustment direction combined with electricity consumption behavior data and operational dynamic information, the objective function and constraints are determined, and a dynamic electricity price generation model is constructed. Using real-time load data and energy storage dispatch results, the electricity price fluctuation range is set, and the situation where the current load exceeds the average value is analyzed to determine the amount of electricity price adjustment.

[0030] Based on the dynamic electricity price generation model, a game theory price adjustment algorithm based on Nash equilibrium is used to iteratively calculate the responses of both parties to the electricity price adjustment amount, and then derive the electricity price adjustment plan.

[0031] Optionally, the user behavior response dataset, generated by combining the real-time electricity price adjustment scheme with the user's economic interest protection needs, is generated through a pre-built user behavior response model, including:

[0032] By using a pre-established electricity price guidance database, we can obtain the specific parameters of the real-time electricity price adjustment plan and historical user feedback data; and classify users according to their economic interests to obtain a preliminary user classification set.

[0033] Based on historical user feedback data, a logistic regression algorithm is used to predict the distribution of user acceptance of electricity price adjustments, and the prediction results of acceptance are determined.

[0034] Based on the acceptance prediction results, the likelihood of changes in user behavior is analyzed to obtain the probability distribution of behavior changes.

[0035] By using the probability distribution of behavioral changes and combining historical data of power grid load, we can simulate the impact trend of user behavior changes on power grid load and determine the key time periods and regional distribution of load changes.

[0036] For key time periods and regional distributions of load changes, obtain user load change values, construct a correlation mapping between user behavior and load changes, and determine the user behavior response dataset.

[0037] Optionally, the step of calculating the volatility of the grid load change trend based on the user behavior response dataset combined with real-time monitoring data of the grid operation status using a grid supply and demand balance dynamic assessment algorithm; and determining the final energy storage dispatch scheme based on whether the volatility of the grid load change trend reaches a preset stability threshold, including:

[0038] Step 1: Use time series analysis to analyze the data in the user behavior response dataset, extract key features of load changes in power grid operation, and determine the periodicity and distribution of abnormal points of load changes.

[0039] Step II: Based on the periodicity of load changes and the distribution of anomalies, apply the dynamic assessment mechanism of supply and demand balance to calculate the volatility of the power grid load change trend.

[0040] Step III: Determine whether the volatility of the power grid load change trend is abnormal. If it is not abnormal, the energy storage resource allocation efficiency optimization scheme shall be adopted as the final energy storage scheduling scheme. Otherwise, adjust the energy storage resource allocation efficiency optimization scheme, recalculate the user behavior response dataset, and return to Step I.

[0041] Furthermore, this invention also provides a control system for an energy storage system based on time-of-use pricing and multimodal optimization, comprising:

[0042] The simulation module is used to simulate supply and demand fluctuations in complex scenarios based on a pre-built energy storage system control strategy model, dynamically optimize the charging and discharging plans of energy storage devices, and generate energy storage resource allocation efficiency optimization schemes.

[0043] The electricity price adjustment module is used to generate real-time electricity price adjustment schemes based on the energy storage resource allocation efficiency optimization scheme, the real-time demand for grid supply and demand balance, changes in user electricity consumption behavior, and the dynamic status of grid operation, using a game theory price adjustment algorithm.

[0044] The behavior response module is used to generate a user behavior response dataset based on the real-time electricity price adjustment scheme and the user's economic interest protection needs, through a pre-built user behavior response model.

[0045] The scheme determination module is used to calculate the volatility of the grid load change trend based on the user behavior response dataset and real-time monitoring data of the grid operation status, using a grid supply and demand balance dynamic evaluation algorithm; and determine the final energy storage dispatch scheme based on whether the volatility of the grid load change trend reaches a preset stability threshold.

[0046] Optionally, the simulation module is specifically used for:

[0047] The pre-built energy storage system control strategy model is used to simulate the operation state under complex scenarios and determine the operation constraints of energy storage devices in different time periods.

[0048] Based on the analysis of simulation results, a control strategy framework is constructed to meet the needs of dynamic adjustment.

[0049] To obtain the optimal response mode of energy storage equipment under supply and demand fluctuations, and to obtain a preliminary scheme for charging and discharging;

[0050] Based on whether the preliminary scheme exceeds the available capacity limit of the energy storage device, determine whether to adjust the parameters of the control strategy framework to obtain an updated scheme that meets the available capacity limit;

[0051] Based on the updated scheme and real-time data of power grid operation, short-term changes in supply and demand fluctuations are monitored. If abnormal fluctuations are detected, the parameters of the control strategy framework are adjusted to optimize the updated scheme and obtain an optimized scheme for energy storage resource allocation efficiency. Otherwise, the updated scheme is directly used as the optimized scheme for energy storage resource allocation efficiency.

[0052] Optionally, a model building module is also included for:

[0053] Data on power grid operation status, energy storage device operation, and user electricity consumption behavior are acquired, and the data on power grid operation status, energy storage device operation, and user electricity consumption behavior are cleaned, standardized, and formatted to obtain a power grid operation dataset.

[0054] Data fusion technology is used to fuse the data in the power grid operation dataset, and a deep reinforcement learning algorithm is used to extract the time-series correlation features of multi-dimensional information to generate a power grid operation correlation feature matrix.

[0055] By utilizing the long short-term memory network's ability to capture long-term dependencies on time-series data, the correlation feature matrix of power grid operation is used to predict the future trends of power grid operation status and user electricity consumption behavior, generate the power grid supply and demand balance demand value and the available capacity range of energy storage devices within the prediction period, and form the power grid operation prediction demand distribution.

[0056] Based on the distribution of demand predicted by power grid operation and the available capacity range of energy storage devices, a control strategy model for energy storage systems based on deep reinforcement learning is constructed.

[0057] Optionally, the electricity price adjustment module includes:

[0058] The dynamic adjustment unit is used to construct a dynamic electricity price generation model based on the energy storage resource allocation efficiency optimization scheme and the real-time demand of power grid supply and demand balance. It adopts a game theory price adjustment algorithm based on Nash equilibrium, combined with user electricity consumption behavior, power grid operation dynamics, and power grid supply and demand balance status, to generate a dynamic electricity price adjustment scheme.

[0059] The linkage unit is used to adjust the energy storage resource allocation efficiency optimization scheme based on the price change trend of the dynamic electricity price adjustment scheme, and determine the resource allocation and electricity price linkage scheme.

[0060] The monitoring and adjustment unit is used to update the parameters in the dynamic electricity price generation model through the resource allocation and electricity price linkage scheme, monitor the real-time demand for power grid supply and demand balance based on the updated dynamic electricity price generation model, and generate a real-time electricity price adjustment scheme.

[0061] Optionally, the dynamic adjustment unit is specifically used for:

[0062] Acquire power grid supply and demand data and balance status information, collect electricity load data and energy storage resource status data in the current period, perform data cleaning and formatting on the electricity load data and energy storage resource status data to obtain a standardized supply and demand dataset;

[0063] The matching degree between the energy storage resource allocation efficiency and the power grid supply and demand data is analyzed based on the supply and demand dataset. When the matching degree is lower than the preset threshold, the allocation efficiency optimization process is triggered. By comparing the operation dynamics in historical data, the adjustment direction of the current energy storage resources is determined.

[0064] Based on the adjustment direction combined with electricity consumption behavior data and operational dynamic information, the objective function and constraints are determined, and a dynamic electricity price generation model is constructed. Using real-time load data and energy storage dispatch results, the electricity price fluctuation range is set, and the situation where the current load exceeds the average value is analyzed to determine the amount of electricity price adjustment.

[0065] Based on the dynamic electricity price generation model, a game theory price adjustment algorithm based on Nash equilibrium is used to iteratively calculate the responses of both parties to the electricity price adjustment amount, and then derive the electricity price adjustment plan.

[0066] Optionally, the behavior response module is specifically used for:

[0067] By using a pre-established electricity price guidance database, we can obtain the specific parameters of the real-time electricity price adjustment plan and historical user feedback data; and classify users according to their economic interests to obtain a preliminary user classification set.

[0068] Based on historical user feedback data, a logistic regression algorithm is used to predict the distribution of user acceptance of electricity price adjustments, and the prediction results of acceptance are determined.

[0069] Based on the acceptance prediction results, the likelihood of changes in user behavior is analyzed to obtain the probability distribution of behavior changes.

[0070] By using the probability distribution of behavioral changes and combining historical data of power grid load, we can simulate the impact trend of user behavior changes on power grid load and determine the key time periods and regional distribution of load changes.

[0071] For key time periods and regional distributions of load changes, obtain user load change values, construct a correlation mapping between user behavior and load changes, and determine the user behavior response dataset.

[0072] Optionally, the scheme determination module is specifically used for:

[0073] Step 1: Use time series analysis to analyze the data in the user behavior response dataset, extract key features of load changes in power grid operation, and determine the periodicity and distribution of abnormal points of load changes.

[0074] Step II: Based on the periodicity of load changes and the distribution of anomalies, apply the dynamic assessment mechanism of supply and demand balance to calculate the volatility of the power grid load change trend.

[0075] Step III: Determine whether the volatility of the power grid load change trend is abnormal. If it is not abnormal, the energy storage resource allocation efficiency optimization scheme shall be adopted as the final energy storage scheduling scheme. Otherwise, adjust the energy storage resource allocation efficiency optimization scheme, recalculate the user behavior response dataset, and return to Step I.

[0076] In another aspect, this application also provides an electronic device, comprising: at least one processor and a memory; the memory and the processor are connected via a bus;

[0077] The memory is used to store one or more programs;

[0078] When the one or more programs are executed by the at least one processor, a control method for an energy storage system based on time-of-use pricing and multimodal optimization, as described above, is implemented.

[0079] Furthermore, this application also provides a readable storage medium having an executable program stored thereon, which, when executed, implements the energy storage system control method based on time-of-use pricing and multimodal optimization as described above.

[0080] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0081] This invention provides a control method for energy storage systems based on time-of-use pricing and multimodal optimization. The method includes: simulating supply and demand fluctuations under complex scenarios using a pre-built energy storage system control strategy model; dynamically optimizing the charging and discharging plans of energy storage devices; and generating an energy storage resource allocation efficiency optimization scheme. Based on the energy storage resource allocation efficiency optimization scheme, real-time demand for grid supply and demand balance, changes in user electricity consumption behavior, and dynamic grid operation status, a game-theoretic price adjustment algorithm is used to generate a real-time electricity price adjustment scheme. Based on the real-time electricity price adjustment scheme and the need to protect user economic interests, a user behavior response dataset is generated using a pre-built user behavior response model. Based on the user behavior response dataset and real-time monitoring data of grid operation status, a grid supply and demand balance dynamic evaluation algorithm is used to calculate the volatility of grid load change trends. Finally, based on whether the volatility of the grid load change trends reaches a preset stability threshold, the final energy storage dispatch scheme is determined. This invention simulates supply and demand fluctuations under complex scenarios using an energy storage system control strategy model, dynamically optimizes the charging and discharging plans of energy storage devices, and combines this with a real-time electricity price adjustment scheme to truly guide user behavior and ensure grid security. Attached Figure Description

[0082] Figure 1This is a flowchart of the energy storage system control method based on time-of-use pricing and multimodal optimization according to the present invention.

[0083] Figure 2 This is a schematic diagram of an electronic device structure according to the present invention. Detailed Implementation

[0084] This invention proposes a control method for energy storage systems based on time-of-use pricing and multimodal optimization. The method includes: constructing a deep reinforcement learning-based control strategy model for the energy storage system based on the predicted demand distribution of the power grid operation and the available capacity range of the energy storage devices. This model dynamically optimizes the charging and discharging plans of the energy storage devices by simulating supply and demand fluctuations under complex scenarios, generating an energy storage resource allocation efficiency optimization scheme. For the adjusted electricity price guidance scheme, considering the need to protect users' economic interests, a user behavior response model based on logistic regression is used to predict the distribution of user acceptance of the price adjustment, analyze the trend of power grid load changes after the user behavior adjustment, and generate a user behavior response dataset. Based on the user behavior response dataset and real-time monitoring data of the power grid operation status, a dynamic evaluation algorithm for power grid supply and demand balance is used to calculate the volatility of the power grid load change trend, determine whether a preset stability threshold is met, and finally, an energy storage dispatch scheme is determined.

[0085] To better understand the present invention, the following description, in conjunction with the accompanying drawings and embodiments, will further illustrate the content of the present invention.

[0086] Example 1:

[0087] A control method for energy storage systems based on time-of-use pricing and multimodal optimization, such as... Figure 1 As shown, it includes:

[0088] Step S1: Based on the pre-built energy storage system control strategy model, simulate supply and demand fluctuations under complex scenarios, dynamically optimize the charging and discharging plans of energy storage devices, and generate an energy storage resource allocation efficiency optimization scheme;

[0089] Step S2: Based on the energy storage resource allocation efficiency optimization scheme, the real-time demand for power grid supply and demand balance, changes in user electricity consumption behavior, and the dynamics of power grid operation status, a game theory price adjustment algorithm is used to generate a real-time electricity price adjustment scheme.

[0090] Step S3: Based on the real-time electricity price adjustment scheme and the need to protect users' economic interests, generate a user behavior response dataset through a pre-built user behavior response model;

[0091] Step S4: Based on the user behavior response dataset and real-time monitoring data of the power grid operation status, calculate the volatility of the power grid load change trend through the power grid supply and demand balance dynamic evaluation algorithm; and determine the final energy storage dispatch scheme based on whether the volatility of the power grid load change trend reaches the preset stability threshold.

[0092] The following is a further explanation of each step:

[0093] Before step S1, the following step also includes: constructing an energy storage system control strategy model. The construction process of the energy storage system control strategy model will be further described below:

[0094] The construction of the energy storage system control strategy model includes:

[0095] Data on power grid operation status, energy storage device operation, and user electricity consumption behavior are acquired, and the data on power grid operation status, energy storage device operation, and user electricity consumption behavior are cleaned, standardized, and formatted to obtain a power grid operation dataset.

[0096] Data fusion technology is used to fuse the data in the power grid operation dataset, and a deep reinforcement learning algorithm is used to extract the time-series correlation features of multi-dimensional information to generate a power grid operation correlation feature matrix.

[0097] By utilizing the long short-term memory network's ability to capture long-term dependencies on time-series data, the correlation feature matrix of power grid operation is used to predict the future trends of power grid operation status and user electricity consumption behavior, generate the power grid supply and demand balance demand value and the available capacity range of energy storage devices within the prediction period, and form the power grid operation prediction demand distribution.

[0098] Based on the distribution of demand predicted by power grid operation and the available capacity range of energy storage devices, a control strategy model for energy storage systems based on deep reinforcement learning is constructed.

[0099] Specifically:

[0100] Step 1: By deploying a distributed sensor network, real-time data on grid operation status, energy storage device operation, and user electricity consumption behavior are collected from various nodes of the power grid. A data acquisition framework containing multi-dimensional information such as voltage, current, energy storage capacity, and user load is constructed. A data preprocessing module is used to clean, standardize, and unify the format of the collected heterogeneous data to generate a unified power grid operation dataset.

[0101] Step 2: For a unified power grid operation dataset, data fusion technology is used to integrate multi-dimensional information in real time. Based on the time-series characteristics of power grid operation status, energy storage equipment operation and user electricity consumption behavior, a time series information fusion model based on an autoregressive integral moving average model is constructed to extract the time-series correlation features of multi-dimensional information and generate a power grid operation correlation feature matrix.

[0102] Step 3: For the power grid operation correlation feature matrix, a long short-term memory network model is adopted. Through its ability to capture long-term dependence on time series data, the future changing trends of power grid operation status and user electricity consumption behavior are predicted, generating the power grid supply and demand balance demand value and the available capacity range of energy storage devices within the prediction period, forming the power grid operation prediction demand distribution.

[0103] Step 4: Based on the predicted demand distribution of the power grid operation and the available capacity range of energy storage devices, construct a control strategy model for the energy storage system based on deep reinforcement learning.

[0104] Step S1: Based on a pre-built energy storage system control strategy model, simulate supply and demand fluctuations under complex scenarios, dynamically optimize the charging and discharging plans of energy storage devices, and generate an energy storage resource allocation efficiency optimization scheme, including:

[0105] The pre-built energy storage system control strategy model is used to simulate the operation state under complex scenarios and determine the operation constraints of energy storage devices in different time periods.

[0106] Based on the analysis of simulation results, a control strategy framework is constructed to meet the needs of dynamic adjustment.

[0107] To obtain the optimal response mode of energy storage equipment under supply and demand fluctuations, and to obtain a preliminary scheme for charging and discharging;

[0108] Based on whether the preliminary scheme exceeds the available capacity limit of the energy storage device, determine whether to adjust the parameters of the control strategy framework to obtain an updated scheme that meets the available capacity limit;

[0109] Based on the updated scheme and real-time data of power grid operation, short-term changes in supply and demand fluctuations are monitored. If abnormal fluctuations are detected, the parameters of the control strategy framework are adjusted to optimize the updated scheme and obtain an optimized scheme for energy storage resource allocation efficiency. Otherwise, the updated scheme is directly used as the optimized scheme for energy storage resource allocation efficiency.

[0110] The specific details of step S1 are as follows:

[0111] Based on the predicted demand distribution of the power grid operation and the available capacity range of energy storage devices, a control strategy model for energy storage systems based on deep reinforcement learning is constructed. By simulating supply and demand fluctuations under complex scenarios, the charging and discharging plans of energy storage devices are dynamically optimized, and an energy storage resource allocation efficiency optimization scheme is generated.

[0112] Historical and real-time power grid operation data are acquired, and features are extracted based on predicted demand and distribution characteristics. A pre-established deep reinforcement learning model is used to perform preliminary data processing, yielding preliminary trend analysis results for supply and demand fluctuations. Based on these preliminary trend analysis results and the available capacity range of energy storage devices, a deep reinforcement learning model is used to simulate the operation under complex scenarios, determining the operational constraints of the energy storage devices at different time periods. Through analysis of the simulation results for complex scenarios, a control strategy framework is constructed to address the need for dynamic adjustment, obtaining the optimal response mode of the energy storage devices under supply and demand fluctuations and determining a preliminary charging and discharging plan. If the preliminary charging and discharging plan exceeds the available capacity limit of the energy storage devices during certain time periods, the resource allocation plan is optimized by adjusting the parameters of the control strategy, resulting in an updated plan that meets the constraints. Based on the updated charging and discharging plan and real-time power grid operation data, an information processing stage is used to monitor short-term changes in supply and demand fluctuations. If abnormal fluctuations are detected, the dynamic adjustment mechanism of the control strategy is triggered to determine the final resource allocation plan. By recording the execution effect of the final resource allocation scheme, the system acquires data on the grid operation status and energy storage device response during the execution process to determine if any deviations exist. If the deviation exceeds a preset threshold, the data is fed back to the deep reinforcement learning model for parameter updates. Based on the feedback data and parameter update results, continuous optimization is performed on the long-term trend and distribution characteristics of grid operation to obtain control strategies that better meet actual needs, resulting in a continuous improvement scheme for energy storage resource allocation efficiency.

[0113] Step S2: Based on the energy storage resource allocation efficiency optimization scheme, the real-time demand for grid supply and demand balance, changes in user electricity consumption behavior, and the dynamics of grid operation status, a game theory-based price adjustment algorithm is used to generate a real-time electricity price adjustment scheme, including:

[0114] Based on the energy storage resource allocation efficiency optimization scheme and the real-time demand of power grid supply and demand balance, a dynamic electricity price generation model is constructed. A game theory price adjustment algorithm based on Nash equilibrium is adopted to generate a dynamic electricity price adjustment scheme by combining user electricity consumption behavior, power grid operation dynamics, and power grid supply and demand balance status.

[0115] Based on the price change trend of the aforementioned dynamic electricity price adjustment scheme, the energy storage resource allocation efficiency optimization scheme is adjusted, and the resource allocation and electricity price linkage scheme is determined.

[0116] The parameters in the dynamic electricity price generation model are updated by the resource allocation and electricity price linkage scheme. Based on the updated dynamic electricity price generation model, the real-time demand for power grid supply and demand balance is monitored, and a real-time electricity price adjustment scheme is generated.

[0117] Furthermore, based on the energy storage resource allocation efficiency optimization scheme and combined with the real-time demand of power grid supply and demand balance, a dynamic electricity price generation model is constructed. A game-theoretic price adjustment algorithm based on Nash equilibrium is used, combined with user electricity consumption behavior, power grid operation dynamics, and power grid supply and demand balance status, to generate a dynamic electricity price adjustment scheme, including:

[0118] Acquire power grid supply and demand data and balance status information, collect electricity load data and energy storage resource status data in the current period, perform data cleaning and formatting on the electricity load data and energy storage resource status data to obtain a standardized supply and demand dataset;

[0119] The matching degree between the energy storage resource allocation efficiency and the power grid supply and demand data is analyzed based on the supply and demand dataset. When the matching degree is lower than the preset threshold, the allocation efficiency optimization process is triggered. By comparing the operation dynamics in historical data, the adjustment direction of the current energy storage resources is determined.

[0120] Based on the adjustment direction combined with electricity consumption behavior data and operational dynamic information, the objective function and constraints are determined, and a dynamic electricity price generation model is constructed. Using real-time load data and energy storage dispatch results, the electricity price fluctuation range is set, and the situation where the current load exceeds the average value is analyzed to determine the amount of electricity price adjustment.

[0121] Based on the dynamic electricity price generation model, a game theory price adjustment algorithm based on Nash equilibrium is used to iteratively calculate the responses of both parties to the electricity price adjustment amount, and then derive the electricity price adjustment plan.

[0122] Step S2 specifically includes:

[0123] Step S21: Obtain power grid supply and demand data and balance status information. Collect electricity load data and energy storage resource status data for the current period through a real-time monitoring system. Perform preliminary cleaning and formatting on the collected data to obtain a standardized supply and demand dataset.

[0124] Step S22: Based on the standardized supply and demand dataset, analyze the matching degree between the allocation efficiency of energy storage resources and the power grid supply and demand. If the matching degree is lower than the preset threshold, trigger the allocation efficiency optimization process. By comparing the operation dynamics in historical data, determine the current adjustment direction of energy storage resources.

[0125] Step S23: Based on the determined adjustment direction, and combining electricity consumption behavior data and operational dynamic information, construct a dynamic electricity price generation model. Use the Nash equilibrium method to conduct game analysis on user electricity consumption behavior and grid operation dynamics to obtain a preliminary set of electricity price parameters.

[0126] Step S24: Based on the preliminary set of electricity price parameters and the current balance between power grid supply and demand, perform price adjustment calculations. If the calculation results show that the price fluctuation exceeds the preset range, perform a second calibration on the parameter set and determine whether the calibrated electricity price parameters meet the balance conditions.

[0127] Step S25: Generate a dynamic electricity price adjustment plan using the calibrated electricity price parameters. Analyze the impact of price change trends in the adjustment plan on electricity consumption behavior and obtain impact data as a reference for subsequent optimization.

[0128] Step S26: Adjust the energy storage resource allocation efficiency strategy based on the impact data. If the impact data indicates insufficient user response, fine-tune the energy storage resource allocation ratio to determine the final resource allocation and electricity price linkage scheme.

[0129] Step S27: Update the parameter configuration in the dynamic electricity price generation model through the final resource allocation and electricity price linkage scheme, monitor the updated model in real time, obtain the new round of power grid supply and demand balance data, and cyclically execute the optimization process.

[0130] Step S3: Based on the real-time electricity price adjustment scheme and the need to protect users' economic interests, generate a user behavior response dataset using a pre-built user behavior response model, including:

[0131] By using a pre-established electricity price guidance database, we can obtain the specific parameters of the real-time electricity price adjustment plan and historical user feedback data; and classify users according to their economic interests to obtain a preliminary user classification set.

[0132] Based on historical user feedback data, a logistic regression algorithm is used to predict the distribution of user acceptance of electricity price adjustments, and the prediction results of acceptance are determined.

[0133] Based on the acceptance prediction results, the likelihood of changes in user behavior is analyzed to obtain the probability distribution of behavior changes.

[0134] By using the probability distribution of behavioral changes and combining historical data of power grid load, we can simulate the impact trend of user behavior changes on power grid load and determine the key time periods and regional distribution of load changes.

[0135] For key time periods and regional distributions of load changes, obtain user load change values, construct a correlation mapping between user behavior and load changes, and determine the user behavior response dataset.

[0136] Step S3 specifically includes:

[0137] For real-time electricity price adjustment schemes, if there is a deviation between the scheme and the grid supply-demand balance demand value in the grid operation forecast demand distribution, a feedback control mechanism will be used to compare the electricity price adjustment parameters with the actual operating data of the energy storage equipment, and the deviation correction value will be calculated using the mean square error formula:

[0138] MSE = 1 / nΣ(yi-yi')^2,

[0139] In the formula, yi is the actual value, yi' is the predicted value, and n is the number of data points, which generates the adjusted electricity price guidance scheme.

[0140] Based on the adjusted electricity price guidance scheme and the need to protect users' economic interests, a user behavior response model based on logistic regression is used to predict the distribution of user acceptance of the electricity price adjustment, analyze the trend of power grid load change after the user behavior adjustment, and generate a user behavior response dataset.

[0141] By establishing a pre-existing electricity price guidance database, specific parameters of the adjustment plan and historical user feedback data are obtained. Users are categorized based on their economic interests to obtain a preliminary user classification set. Using a logistic regression model, and combining parameters related to economic interests and user protection, the acceptance level distribution of each group to the electricity price adjustment is predicted, determining the acceptance level prediction results. Based on the acceptance level prediction results, the probability of changes in user behavior is analyzed. If the acceptance level of a certain user group is lower than a preset threshold, the probability of its behavior change is weighted and calculated to obtain a behavior change probability distribution. Using the behavior change probability distribution, combined with historical grid load data, the impact trend of user behavior changes on grid load is simulated to determine the key time periods and regional distributions of load changes. For the key time periods and regional distributions of load changes, response data from relevant users is obtained, and a correlation mapping between user behavior and load changes is constructed to determine the user behavior response dataset. Based on the user behavior response dataset, the behavioral adjustment characteristics of different user groups under electricity price guidance are analyzed, generating a targeted adjustment suggestion dataset to obtain the final guidance optimization parameters. By using the final guided optimization parameters, the adjustment scheme in the electricity price guidance database is updated, and the matching degree between user acceptance and grid load changes is verified cyclically to determine the optimized scheme dataset.

[0142] Step S4: Based on the user behavior response dataset and real-time monitoring data of the power grid operation status, calculate the volatility of the power grid load change trend using a dynamic evaluation algorithm for power grid supply and demand balance; and determine the final energy storage dispatch scheme based on whether the volatility of the power grid load change trend reaches a preset stability threshold, including:

[0143] Step 1: Use time series analysis to analyze the data in the user behavior response dataset, extract key features of load changes in power grid operation, and determine the periodicity and distribution of abnormal points of load changes.

[0144] Step II: Based on the periodicity of load changes and the distribution of anomalies, apply the dynamic assessment mechanism of supply and demand balance to calculate the volatility of the power grid load change trend.

[0145] Step III: Determine whether the volatility of the power grid load change trend is abnormal. If it is not abnormal, the energy storage resource allocation efficiency optimization scheme shall be adopted as the final energy storage scheduling scheme. Otherwise, adjust the energy storage resource allocation efficiency optimization scheme, recalculate the user behavior response dataset, and return to Step I.

[0146] Step S4 specifically includes:

[0147] Raw datasets are obtained from user behavior and response data, combined with grid operation-related status monitoring information, and integrated into a unified structured data format through data cleaning and preprocessing to obtain a preliminary processed comprehensive dataset. For this preliminary processed comprehensive dataset, real-time data is integrated, and time series analysis methods are used to extract key characteristics of load changes in grid operation, determining the periodicity of load changes and the distribution of anomalies. Based on the periodicity of load changes and the distribution of anomalies, a dynamic supply-demand balance assessment mechanism is applied to calculate the specific values ​​of trend fluctuations and determine whether the trend fluctuations exhibit significant anomalies. If the trend fluctuations exhibit significant anomalies, a stability threshold is used for comparison. Through a preset threshold range, the specific range of fluctuation proportions is obtained to determine whether the fluctuation proportions exceed the expected range. Based on the range distribution of fluctuation proportions and the generation logic of the assessment results, a logistic regression model is used to classify and predict the stability of grid operation, obtaining stability classification labels. Based on the stability classification labels, the characteristic data of load changes and trend fluctuations are integrated, and a comprehensive stability assessment basis for grid operation is generated through data mapping and indicator integration. For this comprehensive stability assessment basis, data visualization tools are used to present user behavior, response data, and grid operation status monitoring information in a multi-dimensional manner, yielding the final assessment result data.

[0148] If the grid operation stability assessment results do not reach the preset stability threshold, the charging and discharging plan of the energy storage equipment will be recalculated through the iterative optimization mechanism of the energy storage system control strategy model, combined with the resource allocation efficiency adjustment requirements. An updated energy storage scheduling scheme will be generated and fed back to the dynamic electricity price generation model to start the next round of electricity price adjustment until the stability assessment results meet the threshold or reach the maximum number of iterations limit.

[0149] The power grid operation data acquisition module acquires real-time power grid operation status data and, combined with historical operation records, analyzes the current power grid stability indicators to obtain preliminary stability assessment results. If the preliminary stability assessment results do not reach a preset threshold, the energy storage system control module is triggered, invoking a pre-established control strategy model to calculate the charging and discharging plan of the energy storage devices based on the current power grid operation status data, and determining a preliminary scheduling scheme. Based on the preliminary scheduling scheme and resource allocation data, resource allocation efficiency is analyzed, and efficiency adjustment logic is used to optimize the charging and discharging plan of the energy storage devices, generating an updated scheduling scheme. The dynamic electricity price generation model acquires the updated scheduling scheme data and, combined with the current power grid operation status, calculates the electricity price adjustment parameters to obtain the adjusted electricity price scheme. If, after applying the adjusted electricity price scheme, the power grid operation stability assessment results still do not reach the preset threshold, an iterative optimization mechanism is used to re-invoke the control strategy model, update the charging and discharging plan based on the latest power grid operation data and electricity price adjustment parameters, and generate a new scheduling scheme. Based on the newly generated scheduling scheme, the calculation process of the dynamic electricity price generation model is repeated to acquire the latest electricity price adjustment parameters and determine whether the power grid operation stability meets the preset threshold, until the result meets the requirements or reaches the preset iteration limit. If the stability requirement is not met even after the maximum number of iterations is reached, the backup strategy module is triggered by recording the current power grid operation status and dispatch scheme data, and the preset backup adjustment logic is used to generate the final emergency dispatch scheme.

[0150] Example 2

[0151] This invention provides a control method for an energy storage system based on time-of-use pricing and multimodal optimization. The following detailed description, using specific implementation examples, illustrates this method, which includes:

[0152] Step S101 involves deploying a distributed sensor network to collect real-time power grid operation status data, energy storage device operation data, and user electricity consumption behavior data from various nodes of the power grid. This constructs a data acquisition framework encompassing multi-dimensional information such as voltage, current, energy storage capacity, and user load. A data preprocessing module cleans, standardizes, and unifies the format of the collected heterogeneous data, generating a unified power grid operation dataset. The distributed sensor network collects operation status data, energy storage device information, and electricity consumption behavior records from power grid nodes, constructing a multi-dimensional data framework including voltage data, current data, energy storage capacity, and load information, thus obtaining a preliminary power grid operation dataset. The preprocessing module cleans the heterogeneous data in the preliminary power grid operation dataset. For missing or outlier data, if the voltage or current data of a node exceeds a preset threshold range, interpolation is used to fill in the missing parts, determining the cleaned basic dataset. Based on the cleaned basic dataset, a standardization method is used to unify the dimensions of the voltage data, current data, and energy storage capacity, obtaining a standardized dataset with a consistent format. By extracting features from load information and electricity consumption behavior in a standardized dataset, a support vector machine algorithm is used to classify user electricity consumption patterns and determine the load distribution characteristics in different time periods. Based on the load distribution characteristics and combined with the status and capacity data of energy storage devices, if the energy storage capacity of a node falls below a preset lower limit, an energy storage scheduling mechanism is triggered to obtain an adjusted energy storage allocation scheme. After obtaining the adjusted energy storage allocation scheme, the operating status data of grid nodes is monitored in real time. If abnormal fluctuations in voltage or current data are detected, comparative analysis is performed using historical standardized datasets to determine the specific source of the abnormal fluctuations. Based on the source information of the abnormal fluctuations, combined with the load distribution characteristics and the energy storage allocation scheme, a pre-established grid operation model is used for dynamic adjustment to obtain optimized grid operation parameters.

[0153] Step S102: For a unified power grid operation dataset, data fusion technology is used to integrate multi-dimensional information in real time. Based on the time-series characteristics of power grid operation status, energy storage equipment operation and user electricity consumption behavior, a time-series information fusion model based on an autoregressive integral moving average model is constructed to extract the time-series correlation features of multi-dimensional information and generate a power grid operation correlation feature matrix.

[0154] Step 1: Obtain raw multidimensional information containing operating status, energy storage device data, and electricity consumption behavior from the power grid operation-related dataset. Preprocess this information, cleaning outliers and standardizing the format to obtain a preliminary multidimensional dataset. Step 2: For the preliminary multidimensional dataset, use data fusion technology to integrate the multidimensional information on operating status, energy storage devices, and electricity consumption behavior in real time. Determine the unified data structure after fusion by standardizing data from different sources. Step 3: Based on the unified data structure after fusion, extract time-series characteristic data related to the time series. Use an autoregressive integral moving average model to model the time-series characteristic data, obtaining a time-series correlated feature set. Step 4: Extract a subset of features directly related to the power grid operating status from the time-series correlated feature set. If a feature value in the feature subset deviates from a preset threshold, smooth it to determine the optimized feature subset. Step 5: For the optimized feature subset, combine the time-series characteristics of energy storage devices and electricity consumption behavior to construct a correlation feature matrix for power grid operation. Obtain the final feature matrix structure through matrix recombination technology. Step 6: Based on the final feature matrix structure, analyze the dynamic correlation between the power grid operating status and energy storage devices and electricity consumption behavior. Through dimensionality reduction of the feature matrix, obtain a simplified correlation feature representation. Step 7: For the simplified correlation feature representation, generate a multi-dimensional time-series correlation mapping of power grid operation. By storing and updating the mapping results, determine the real-time characteristic change trend of power grid operation.

[0155] Specifically, for a unified power grid operation dataset, multi-dimensional information is first integrated in real time using data fusion technology.

[0156] For example, suppose we collect data from three dimensions: grid operation status, energy storage device operation data, and user electricity consumption behavior. Specifically, the grid voltage value is collected every minute, the charging and discharging power of energy storage devices is recorded every 5 minutes, and user electricity consumption is statistically analyzed every hour. These data have inconsistent timestamps and different formats. To achieve integration, a time alignment algorithm is used to unify all data to a minute-level time granularity. Missing values ​​are filled using linear interpolation. For example, for the power data of energy storage devices, interpolation is performed within the missing 4 minutes. Assuming the power is 50.2 kW at a certain moment and 52.8 kW at the next moment, the missing intermediate values ​​are calculated linearly as 51.0, 51.5, and 52.0 kW. Subsequently, based on the time-series characteristics of grid operation status, energy storage device operation, and user electricity consumption behavior, a time series information fusion model based on the Autoregressive Integral Moving Average (ARIMA) model is constructed. The specific method is to first perform a stationarity test on the data of each dimension. For example, the ADF test is used to determine whether the grid voltage series is stationary. If the p-value is less than 0.05 (0.03), the series is stationary; otherwise, first-order differencing is performed. Next, ARIMA model parameters are set. For example, an ARIMA(1,1,1) model is fitted to voltage data, where the autoregressive order p = 1, the differencing order d = 1, and the moving average order q = 1. Parameters are optimized by minimizing the AIC value. Residual analysis of the model's predicted and actual values ​​is obtained. The mean residual is close to 0 and the variance is 0.25, indicating a good model fit. Finally, time-series correlation features of multi-dimensional information are extracted to generate a power grid operation correlation feature matrix. For example, by calculating the Pearson correlation coefficients between voltage, power, and electricity consumption, the correlation coefficient between voltage and power is 0.85, and the correlation coefficient between power and electricity consumption is 0.72. A 3×3 correlation matrix is ​​constructed, with matrix elements representing the correlation coefficients between each dimension, used for subsequent analysis of the dynamic correlation of power grid operation. To form a logical chain, the correlation matrix can be further input into the power grid stability prediction model. Combined with historical fault data, the impact of the feature matrix on the fault probability is analyzed. For example, dimensional combinations with correlation coefficients higher than 0.8 require key monitoring, thus achieving a complete closed loop from data fusion to feature extraction to business application.

[0157] Step S103: For the power grid operation correlation feature matrix, a long short-term memory network model is adopted. Through its ability to capture long-term dependence on time series data, the future changing trends of power grid operation status and user electricity consumption behavior are predicted, generating the power grid supply and demand balance demand value and the available capacity range of energy storage devices within the prediction period, thus forming the power grid operation prediction demand distribution.

[0158] For power grid operation data, multi-dimensional correlation features are extracted from historical records to construct an initial feature matrix, covering time-series data and user electricity consumption behavior information. A standardized first feature matrix is ​​obtained through data preprocessing. A Long Short-Term Memory (LSTM) network model is used as input to the first feature matrix to capture the long-term dependency characteristics of the time-series data, analyze the potential patterns of power grid operation status and electricity consumption behavior, and output predicted trend data. Based on the trend data, the supply-demand balance demand value for the predicted time period is calculated. If the calculated result exceeds a preset threshold range, the trend data is recalibrated to obtain an adjusted first demand value. Using the adjusted first demand value and the operating parameters of energy storage devices, the available capacity range is analyzed. If the available capacity is lower than a preset lower limit, a capacity replenishment mechanism is triggered to determine the final second demand value. The second demand value is obtained and combined with real-time power grid operation status data to generate a predicted demand distribution for the predicted time period, outputting the first distribution result. For the first distribution result, a comparative analysis of historical supply-demand balance data is introduced. If the distribution deviation exceeds a preset tolerance range, the first distribution result is corrected to obtain the final second distribution result. Based on the second distribution results, a dynamic adjustment strategy for power grid operation is generated, and a comprehensive scheduling scheme for supply and demand balance and available capacity is output.

[0159] Step S104: Based on the predicted demand distribution of the power grid operation and the available capacity range of the energy storage equipment, construct a control strategy model for the energy storage system based on deep reinforcement learning. By simulating supply and demand fluctuations under complex scenarios, dynamically optimize the charging and discharging plan of the energy storage equipment and generate an energy storage resource allocation efficiency optimization scheme.

[0160] Historical and real-time power grid operation data are acquired, and features are extracted based on predicted demand and distribution characteristics. A pre-established deep reinforcement learning model is used to perform preliminary data processing, yielding preliminary trend analysis results for supply and demand fluctuations. Based on these preliminary trend analysis results and the available capacity range of energy storage devices, a deep reinforcement learning model is used to simulate the operation under complex scenarios, determining the operational constraints of the energy storage devices at different time periods. Through analysis of the simulation results for complex scenarios, a control strategy framework is constructed to address the need for dynamic adjustment, obtaining the optimal response mode of the energy storage devices under supply and demand fluctuations and determining a preliminary charging and discharging plan. If the preliminary charging and discharging plan exceeds the available capacity limit of the energy storage devices during certain time periods, the resource allocation plan is optimized by adjusting the parameters of the control strategy, resulting in an updated plan that meets the constraints. Based on the updated charging and discharging plan and real-time power grid operation data, an information processing stage is used to monitor short-term changes in supply and demand fluctuations. If abnormal fluctuations are detected, the dynamic adjustment mechanism of the control strategy is triggered to determine the final resource allocation plan. By recording the execution effect of the final resource allocation scheme, the system acquires data on the grid operation status and energy storage device response during the execution process to determine if any deviations exist. If the deviation exceeds a preset threshold, the data is fed back to the deep reinforcement learning model for parameter updates. Based on the feedback data and parameter update results, continuous optimization is performed on the long-term trend and distribution characteristics of grid operation to obtain control strategies that better meet actual needs, resulting in a continuous improvement scheme for energy storage resource allocation efficiency.

[0161] Specifically, regarding the predicted demand distribution for power grid operation, load forecasting is first performed using historical data and real-time monitoring, combined with machine learning algorithms such as Long Short-Term Memory (LSTM) networks. Assuming a region's average daily load over the past year is 100,000 kW and peak load is 150,000 kW, the LSTM model predicts load fluctuations for the next 24 hours, indicating a potential peak of 160,000 kW with an error margin controlled within 5%. Analysis shows peak periods are concentrated between 12 PM and 2 PM, providing data support for subsequent energy storage dispatch. Next, considering the available capacity range of energy storage devices, assuming a total energy storage system capacity of 5,000 kWh and an available capacity of 80%, or 4,000 kWh, a capacity allocation algorithm distributes the available capacity according to the predicted load proportion. Of this, 3,000 kWh is used for peak shaving, and 1,000 kWh is reserved as reserve capacity. The system analyzes whether the capacity meets demand; if insufficient, it automatically calls upon backup power resources through the system interface to ensure supply and demand balance. Subsequently, a control strategy model for the energy storage system based on deep reinforcement learning was constructed. The Deep Deterministic Policy Gradient (DDPG) algorithm was adopted, and the reward function was set as a weighted sum of grid stability and economic benefits, with weights of 0.6 and 0.4, respectively. By simulating supply and demand fluctuation scenarios (such as a sudden 10% load increase), the charging and discharging schedule was dynamically adjusted. The optimal charging time was determined to be from 2:00 AM to 5:00 AM with a power output of 1000 kW, and the optimal discharging time was from 12:00 PM to 2:00 PM with a power output of 1500 kW. After optimization, the grid voltage fluctuation rate decreased from 5% to 2%. Finally, an optimized energy storage resource allocation efficiency scheme was generated. Based on the above results, the system automatically generated scheduling instructions to prioritize the allocation of energy storage resources during peak hours. Calculations showed that resource utilization increased to 90%, and economic benefits improved by approximately 15%. The comparison before and after optimization was displayed through a data visualization interface to ensure the traceability of the scheme and form a complete closed-loop logic from prediction to optimization.

[0162] Step S105: For the energy storage resource allocation efficiency optimization scheme, combined with the real-time demand of power grid supply and demand balance, a dynamic electricity price generation model is constructed. A game theory price adjustment algorithm based on Nash equilibrium is adopted. By comprehensively considering changes in user electricity consumption behavior and the dynamics of power grid operation status, the electricity price adjustment parameters adapted to the current scenario are calculated, and a real-time electricity price adjustment scheme is generated.

[0163] Step 1: Obtain power grid supply and demand data and balance status information. Collect electricity load data and energy storage resource status data for the current period through a real-time monitoring system. Perform preliminary cleaning and formatting on the collected data to obtain a standardized supply and demand dataset.

[0164] Step 2: Based on the standardized supply and demand dataset, analyze the matching degree between the energy storage resource allocation efficiency and the grid supply and demand. If the matching degree is lower than the preset threshold, trigger the allocation efficiency optimization process. By comparing the operation dynamics in historical data, determine the current adjustment direction of energy storage resources.

[0165] Step 3: Based on the determined adjustment direction, and combining electricity consumption behavior data and operational dynamic information, construct a dynamic electricity price generation model. Use the Nash equilibrium method to conduct game analysis on user electricity consumption behavior and grid operation dynamics to obtain a preliminary set of electricity price parameters.

[0166] Step 4: Based on the preliminary set of electricity price parameters and the current balance between power grid supply and demand, perform price adjustment calculations. If the calculation results show that the price fluctuation exceeds the preset range, perform a second calibration of the parameter set and determine whether the calibrated electricity price parameters meet the balance conditions.

[0167] Step 5: Generate a dynamic electricity price adjustment plan using the calibrated electricity price parameters. Analyze the impact of price change trends in the adjustment plan on electricity consumption behavior and obtain impact data as a reference for subsequent optimization.

[0168] Step 6: Adjust the energy storage resource allocation efficiency strategy based on the impact data. If the impact data indicates insufficient user response, fine-tune the energy storage resource allocation ratio to determine the final resource allocation and electricity price linkage scheme.

[0169] Step 7: Update the parameter configuration in the dynamic electricity price generation model through the final resource allocation and electricity price linkage scheme, monitor the updated model in real time, obtain the new round of power grid supply and demand balance data, and cyclically execute the optimization process.

[0170] Specifically, regarding the implementation of the energy storage resource allocation efficiency optimization scheme, firstly, real-time supply and demand data are acquired through the smart grid data acquisition system. Assuming the current peak grid load is 500 MW, the available capacity of the energy storage system is 800 MW, and the system detects a supply-demand gap of 300 MW during peak load periods, the optimization algorithm employs linear programming. The objective function is set as minimizing energy storage dispatch costs, with the constraint that energy storage discharge power does not exceed 200 MW / hour. The calculation yields the optimal dispatch scheme as releasing 150 MW of energy storage power per hour, ensuring supply and demand balance while reducing dispatch costs by approximately 10%. Next, combining the real-time demand for grid supply and demand balance, a dynamic electricity price generation model is constructed. Using real-time load data and energy storage dispatch results, the electricity price fluctuation range is set to 0.5 to 1.5 yuan / kWh, with an initial price of 1.0 yuan / kWh. Analysis of the current load exceeding the average by 20% determines that the electricity price needs to be increased by 15%. Subsequently, a game-theoretic price adjustment algorithm based on Nash equilibrium was adopted. Assuming the user group is divided into industrial and residential categories, with industrial users having a price sensitivity of 0.3 and residential users 0.5, the algorithm iteratively calculates the responses of both sides to price adjustments, arriving at an equilibrium point where the price is increased to 1.2 yuan / kWh. At this point, user electricity consumption decreases by approximately 8%, and the grid load drops to 4700 MW, approaching equilibrium. Further, considering changes in user electricity consumption behavior and the dynamic state of the grid operation, the system collects real-time user electricity consumption data, finding that industrial users reduce consumption by 100 MW and residential users by 50 MW. Combining this with grid operation status, such as a frequency deviation of 0.02 Hz, the electricity price adjustment parameters are calculated, concluding that a slight adjustment to 1.18 yuan / kWh is needed to maintain stability. Finally, a real-time electricity price adjustment plan is generated. The system automatically pushes the adjusted electricity price to the user terminal through smart meters, while recording the adjustment effect. Analysis shows that the load has been further reduced to 4650 MW, the supply and demand gap has been narrowed to 50 MW, and the optimization effect is significant. The whole process achieves closed-loop control through automated algorithms and data analysis.

[0171] Step S106: For the real-time electricity price adjustment scheme, if there is a deviation between it and the grid supply and demand balance demand value in the grid operation forecast demand distribution, the electricity price adjustment parameters are compared with the actual operation data of the energy storage device through the feedback control mechanism, and the deviation correction value is calculated using the mean square error. The mean square error formula is MSE=1 / nΣ(yi-yi')^2, where yi is the actual value, yi' is the forecast value, and n is the number of data points, thereby generating the adjusted electricity price guidance scheme.

[0172] Step S107: Based on the adjusted electricity price guidance scheme and combined with the user's need to protect economic interests, a user behavior response model based on logistic regression is used to predict the distribution of user acceptance of the electricity price adjustment, analyze the trend of power grid load change after user behavior adjustment, and generate a user behavior response dataset.

[0173] By establishing a pre-existing electricity price guidance database, specific parameters of the adjustment plan and historical user feedback data are obtained. Users are categorized based on their economic interests to obtain a preliminary user classification set. Using a logistic regression model, and combining parameters related to economic interests and user protection, the acceptance level distribution of each group to the electricity price adjustment is predicted, determining the acceptance level prediction results. Based on the acceptance level prediction results, the probability of changes in user behavior is analyzed. If the acceptance level of a certain user group is lower than a preset threshold, the probability of its behavior change is weighted and calculated to obtain a behavior change probability distribution. Using the behavior change probability distribution, combined with historical grid load data, the impact trend of user behavior changes on grid load is simulated to determine the key time periods and regional distributions of load changes. For the key time periods and regional distributions of load changes, response data from relevant users is obtained, and a correlation mapping between user behavior and load changes is constructed to determine the user behavior response dataset. Based on the user behavior response dataset, the behavioral adjustment characteristics of different user groups under electricity price guidance are analyzed, generating a targeted adjustment suggestion dataset to obtain the final guidance optimization parameters. By using the final guided optimization parameters, the adjustment scheme in the electricity price guidance database is updated, and the matching degree between user acceptance and grid load changes is verified cyclically to determine the optimized scheme dataset.

[0174] Specifically, regarding the adjusted electricity price guidance plan, and considering the need to protect users' economic interests, the first step is to construct a user behavior response model based on logistic regression to predict the distribution of user acceptance of the price adjustment. Assuming the electricity price is adjusted from 0.5 yuan per kilowatt-hour to 0.6 yuan, an increase of 20%, we collected electricity consumption data from 10,000 users, including monthly average electricity consumption, income level, and sensitivity to electricity prices. Using a logistic regression model, the probability of users accepting the price adjustment is used as the dependent variable, and the characteristic variables are input into the model. Maximum likelihood estimation is used for parameter fitting, resulting in a model prediction accuracy of 85%. The results show that approximately 60% of users have a high acceptance rate, while 40% of users may reduce their electricity consumption due to economic pressure. Next, we analyzed the trend of grid load changes after user behavior adjustments. Assuming users with low acceptance levels reduce their electricity consumption by 10%, from an average of 300 kWh per month to 270 kWh, while users with high acceptance levels maintain their current consumption, we predicted an overall grid load decrease of approximately 4%, from 3 million kWh to 2.88 million kWh, through a weighted calculation of all users' electricity consumption changes. To further verify the impact of load changes on grid stability, we introduced peak load analysis. Assuming the peak load accounts for 30% of the total load, and the adjusted peak load decreases to 28.8%, time series analysis predicts that the grid load fluctuation range within the next month will be within ±2%, indicating that the impact of electricity price adjustments on grid stability is controllable. Finally, we generated a user behavior response dataset, storing user IDs, electricity consumption changes, acceptance probabilities, and other information as a structured data table containing 10,000 records. Each record's fields include user category (high acceptance / low acceptance) and adjusted electricity consumption (accurate to 1 kWh). The dataset was automatically generated and stored through a database management system and can be used for subsequent grid dispatch optimization. Through the above process, from user behavior prediction to grid load analysis and then to data generation, a complete logical chain is formed to ensure that the electricity price adjustment plan protects the economic interests of users while maintaining the stability of grid operation. Step S108: Based on the user behavior response dataset and combined with real-time monitoring data of grid operation status, the volatility of grid load change trend is calculated using a grid supply and demand balance dynamic assessment algorithm to determine whether it meets the preset stability threshold. A volatility of less than 5% is considered stable, and a grid operation stability assessment result is generated.

[0175] Raw datasets are obtained from user behavior and response data, combined with grid operation-related status monitoring information, and integrated into a unified structured data format through data cleaning and preprocessing to obtain a preliminary processed comprehensive dataset. For this preliminary processed comprehensive dataset, real-time data is integrated, and time series analysis methods are used to extract key characteristics of load changes in grid operation, determining the periodicity of load changes and the distribution of anomalies. Based on the periodicity of load changes and the distribution of anomalies, a dynamic supply-demand balance assessment mechanism is applied to calculate the specific values ​​of trend fluctuations and determine whether the trend fluctuations exhibit significant anomalies. If the trend fluctuations exhibit significant anomalies, a stability threshold is used for comparison. Through a preset threshold range, the specific range of fluctuation proportions is obtained to determine whether the fluctuation proportions exceed the expected range. Based on the range distribution of fluctuation proportions and the generation logic of the assessment results, a logistic regression model is used to classify and predict the stability of grid operation, obtaining stability classification labels. Based on the stability classification labels, the characteristic data of load changes and trend fluctuations are integrated, and a comprehensive stability assessment basis for grid operation is generated through data mapping and indicator integration. For this comprehensive stability assessment basis, data visualization tools are used to present user behavior, response data, and grid operation status monitoring information in a multi-dimensional manner, yielding the final assessment result data.

[0176] Specifically, regarding the combination of user behavior response datasets and real-time monitoring data of power grid operation status, user electricity consumption behavior data is first acquired through a data acquisition system. For example, the electricity consumption data of 1,000 households in a certain area over 24 hours, with an average consumption of 5.2 kWh per household. Simultaneously, real-time load data is acquired from the power grid monitoring system, such as the current total power grid load of 5,200 MW and the peak load of 5,500 MW. Next, a dynamic assessment algorithm for power grid supply and demand balance is used to calculate the volatility of power grid load change trends. The specific algorithm is as follows: based on the hourly load change percentage, the standard deviation of the load data over the past 12 hours is calculated. Assuming the load change data is [5100, 5150, 5200, 5250, 5300, 5200, 5150, 5100, 5050, 5100, 5150, 5200] MW, the calculated standard deviation is 78.6, and the volatility is 1.51% (i.e., the standard deviation divided by the average load of 5,200 MW). Subsequently, the volatility was compared with a preset stability threshold of 5%. Finding that 1.51% was less than 5%, the power grid was judged to be operating stably. Finally, a power grid stability assessment result was generated, and the system automatically output a report showing a volatility of 1.51% and a "stable" status. The result was stored in the database, and simultaneously, downstream business systems were linked. If the volatility exceeded the threshold, load regulation strategies were automatically triggered, such as calling backup power or adjusting user-side demand response plans, to ensure the safe operation of the power grid. Through this process, from data acquisition to algorithm calculation to result output, a complete logical chain was formed, ensuring the scientific rigor and real-time nature of the assessment.

[0177] Step S109: If the grid operation stability assessment results do not reach the preset stability threshold, the charging and discharging plan of the energy storage equipment is recalculated through the iterative optimization mechanism of the energy storage system control strategy model, combined with the resource allocation efficiency adjustment requirements, to generate an updated energy storage scheduling scheme. This scheme is then fed back to the dynamic electricity price generation model to initiate the next round of electricity price adjustment until the stability assessment results meet the threshold or reach the maximum iteration limit.

[0178] The power grid operation data acquisition module acquires real-time power grid operation status data and, combined with historical operation records, analyzes the current power grid stability indicators to obtain preliminary stability assessment results. If the preliminary stability assessment results do not reach a preset threshold, the energy storage system control module is triggered, invoking a pre-established control strategy model to calculate the charging and discharging plan of the energy storage devices based on the current power grid operation status data, and determining a preliminary scheduling scheme. Based on the preliminary scheduling scheme and resource allocation data, resource allocation efficiency is analyzed, and efficiency adjustment logic is used to optimize the charging and discharging plan of the energy storage devices, generating an updated scheduling scheme. The dynamic electricity price generation model acquires the updated scheduling scheme data and, combined with the current power grid operation status, calculates the electricity price adjustment parameters to obtain the adjusted electricity price scheme. If, after applying the adjusted electricity price scheme, the power grid operation stability assessment results still do not reach the preset threshold, an iterative optimization mechanism is used to re-invoke the control strategy model, update the charging and discharging plan based on the latest power grid operation data and electricity price adjustment parameters, and generate a new scheduling scheme. Based on the newly generated scheduling scheme, the calculation process of the dynamic electricity price generation model is repeated to acquire the latest electricity price adjustment parameters and determine whether the power grid operation stability meets the preset threshold, until the result meets the requirements or reaches the preset iteration limit. If the stability requirement is not met even after the maximum number of iterations is reached, the backup strategy module is triggered by recording the current power grid operation status and dispatch scheme data, and the preset backup adjustment logic is used to generate the final emergency dispatch scheme.

[0179] Example 3

[0180] Based on the same inventive concept, this invention also provides a control system for an energy storage system based on time-of-use pricing and multimodal optimization, comprising:

[0181] The simulation module is used to simulate supply and demand fluctuations in complex scenarios based on a pre-built energy storage system control strategy model, dynamically optimize the charging and discharging plans of energy storage devices, and generate energy storage resource allocation efficiency optimization schemes.

[0182] The electricity price adjustment module is used to generate real-time electricity price adjustment schemes based on the energy storage resource allocation efficiency optimization scheme, the real-time demand for grid supply and demand balance, changes in user electricity consumption behavior, and the dynamic status of grid operation, using a game theory price adjustment algorithm.

[0183] The behavior response module is used to generate a user behavior response dataset based on the real-time electricity price adjustment scheme and the user's economic interest protection needs, through a pre-built user behavior response model.

[0184] The scheme determination module is used to calculate the volatility of the grid load change trend based on the user behavior response dataset and real-time monitoring data of the grid operation status, using a grid supply and demand balance dynamic evaluation algorithm; and determine the final energy storage dispatch scheme based on whether the volatility of the grid load change trend reaches a preset stability threshold.

[0185] Optionally, the simulation module is specifically used for:

[0186] The pre-built energy storage system control strategy model is used to simulate the operation state under complex scenarios and determine the operation constraints of energy storage devices in different time periods.

[0187] Based on the analysis of simulation results, a control strategy framework is constructed to meet the needs of dynamic adjustment.

[0188] To obtain the optimal response mode of energy storage equipment under supply and demand fluctuations, and to obtain a preliminary scheme for charging and discharging;

[0189] Based on whether the preliminary scheme exceeds the available capacity limit of the energy storage device, determine whether to adjust the parameters of the control strategy framework to obtain an updated scheme that meets the available capacity limit;

[0190] Based on the updated scheme and real-time data of power grid operation, short-term changes in supply and demand fluctuations are monitored. If abnormal fluctuations are detected, the parameters of the control strategy framework are adjusted to optimize the updated scheme and obtain an optimized scheme for energy storage resource allocation efficiency. Otherwise, the updated scheme is directly used as the optimized scheme for energy storage resource allocation efficiency.

[0191] Optionally, a model building module is also included for:

[0192] Data on power grid operation status, energy storage device operation, and user electricity consumption behavior are acquired, and the data on power grid operation status, energy storage device operation, and user electricity consumption behavior are cleaned, standardized, and formatted to obtain a power grid operation dataset.

[0193] Data fusion technology is used to fuse the data in the power grid operation dataset, and a deep reinforcement learning algorithm is used to extract the time-series correlation features of multi-dimensional information to generate a power grid operation correlation feature matrix.

[0194] By utilizing the long short-term memory network's ability to capture long-term dependencies on time-series data, the correlation feature matrix of power grid operation is used to predict the future trends of power grid operation status and user electricity consumption behavior, generate the power grid supply and demand balance demand value and the available capacity range of energy storage devices within the prediction period, and form the power grid operation prediction demand distribution.

[0195] Based on the distribution of demand predicted by power grid operation and the available capacity range of energy storage devices, a control strategy model for energy storage systems based on deep reinforcement learning is constructed.

[0196] Optionally, the electricity price adjustment module includes:

[0197] The dynamic adjustment unit is used to construct a dynamic electricity price generation model based on the energy storage resource allocation efficiency optimization scheme and the real-time demand of power grid supply and demand balance. It adopts a game theory price adjustment algorithm based on Nash equilibrium, combined with user electricity consumption behavior, power grid operation dynamics, and power grid supply and demand balance status, to generate a dynamic electricity price adjustment scheme.

[0198] The linkage unit is used to adjust the energy storage resource allocation efficiency optimization scheme based on the price change trend of the dynamic electricity price adjustment scheme, and determine the resource allocation and electricity price linkage scheme.

[0199] The monitoring and adjustment unit is used to update the parameters in the dynamic electricity price generation model through the resource allocation and electricity price linkage scheme, monitor the real-time demand for power grid supply and demand balance based on the updated dynamic electricity price generation model, and generate a real-time electricity price adjustment scheme.

[0200] Optionally, the dynamic adjustment unit is specifically used for:

[0201] Acquire power grid supply and demand data and balance status information, collect electricity load data and energy storage resource status data in the current period, perform data cleaning and formatting on the electricity load data and energy storage resource status data to obtain a standardized supply and demand dataset;

[0202] The matching degree between the energy storage resource allocation efficiency and the power grid supply and demand data is analyzed based on the supply and demand dataset. When the matching degree is lower than the preset threshold, the allocation efficiency optimization process is triggered. By comparing the operation dynamics in historical data, the adjustment direction of the current energy storage resources is determined.

[0203] Based on the adjustment direction combined with electricity consumption behavior data and operational dynamic information, the objective function and constraints are determined, and a dynamic electricity price generation model is constructed. Using real-time load data and energy storage dispatch results, the electricity price fluctuation range is set, and the situation where the current load exceeds the average value is analyzed to determine the amount of electricity price adjustment.

[0204] Based on the dynamic electricity price generation model, a game theory price adjustment algorithm based on Nash equilibrium is used to iteratively calculate the responses of both parties to the electricity price adjustment amount, and then derive the electricity price adjustment plan.

[0205] Optionally, the behavior response module is specifically used for:

[0206] By using a pre-established electricity price guidance database, we can obtain the specific parameters of the real-time electricity price adjustment plan and historical user feedback data; and classify users according to their economic interests to obtain a preliminary user classification set.

[0207] Based on historical user feedback data, a logistic regression algorithm is used to predict the distribution of user acceptance of electricity price adjustments, and the prediction results of acceptance are determined.

[0208] Based on the acceptance prediction results, the likelihood of changes in user behavior is analyzed to obtain the probability distribution of behavior changes.

[0209] By using the probability distribution of behavioral changes and combining historical data of power grid load, we can simulate the impact trend of user behavior changes on power grid load and determine the key time periods and regional distribution of load changes.

[0210] For key time periods and regional distributions of load changes, obtain user load change values, construct a correlation mapping between user behavior and load changes, and determine the user behavior response dataset.

[0211] Optionally, the scheme determination module is specifically used for:

[0212] Step 1: Use time series analysis to analyze the data in the user behavior response dataset, extract key features of load changes in power grid operation, and determine the periodicity and distribution of abnormal points of load changes.

[0213] Step II: Based on the periodicity of load changes and the distribution of anomalies, apply the dynamic assessment mechanism of supply and demand balance to calculate the volatility of the power grid load change trend.

[0214] Step III: Determine whether the volatility of the power grid load change trend is abnormal. If it is not abnormal, the energy storage resource allocation efficiency optimization scheme shall be adopted as the final energy storage scheduling scheme. Otherwise, adjust the energy storage resource allocation efficiency optimization scheme, recalculate the user behavior response dataset, and return to Step I.

[0215] Example 4

[0216] like Figure 2 As shown, the present invention also provides an electronic device, which may be a computer device, a microcontroller device, a smart mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, processor, and transceiver component are connected via a bus; the memory can be used to store executable programs, and an exemplary executable program may include instructions; the processor is used to execute the instructions stored in the memory. The memory can also be used to store data, which can be accessed and / or modified when instructions are executed.

[0217] The processor may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and it is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the storage medium to realize the corresponding method flow or corresponding function, so as to realize the steps of the energy storage system control method based on time-of-use pricing and multimodal optimization in the above embodiments.

[0218] Example 4

[0219] Based on the same inventive concept, this invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory). This readable storage medium is a memory device within an electronic device used to store programs and data. It is understood that the storage medium here can include both built-in storage media within the electronic device and extended storage media supported by the electronic device. The storage medium provides storage space, which stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more executable programs (including program code). It should be noted that the storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. Loading and executing one or more instructions stored in the storage medium by the processor can implement the steps of the energy storage system control method based on time-of-use pricing and multimodal optimization in the above embodiments.

[0220] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0221] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0222] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0223] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0224] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of the claims of the present invention pending approval.

Claims

1. A control method for an energy storage system based on time-of-use pricing and multimodal optimization, characterized in that, include: Based on a pre-built energy storage system control strategy model, the system simulates supply and demand fluctuations under complex scenarios, dynamically optimizes the charging and discharging plans of energy storage devices, and generates an energy storage resource allocation efficiency optimization scheme. Based on the energy storage resource allocation efficiency optimization scheme, the real-time demand for power grid supply and demand balance, changes in user electricity consumption behavior and the dynamics of power grid operation status, a game theory price adjustment algorithm is used to generate a real-time electricity price adjustment scheme. Based on the real-time electricity price adjustment scheme and the need to protect users' economic interests, a user behavior response dataset is generated through a pre-built user behavior response model. Based on user behavior response datasets and real-time monitoring data of power grid operation status, the volatility of power grid load change trends is calculated using a dynamic evaluation algorithm for power grid supply and demand balance. Based on whether the volatility of the power grid load change trend reaches a preset stability threshold, the final energy storage scheduling scheme is determined.

2. The method as described in claim 1, characterized in that, The construction of the energy storage system control strategy model includes: Data on power grid operation status, energy storage device operation, and user electricity consumption behavior are acquired, and the data on power grid operation status, energy storage device operation, and user electricity consumption behavior are cleaned, standardized, and formatted to obtain a power grid operation dataset. Data fusion technology is used to fuse the data in the power grid operation dataset, and a deep reinforcement learning algorithm is used to extract the time-series correlation features of multi-dimensional information to generate a power grid operation correlation feature matrix. By utilizing the long short-term memory network's ability to capture long-term dependencies on time-series data, the correlation feature matrix of power grid operation is used to predict the future trends of power grid operation status and user electricity consumption behavior, generate the power grid supply and demand balance demand value and the available capacity range of energy storage devices within the prediction period, and form the power grid operation prediction demand distribution. Based on the distribution of demand predicted by power grid operation and the available capacity range of energy storage devices, a control strategy model for energy storage systems based on deep reinforcement learning is constructed.

3. The method as described in claim 1, characterized in that, The pre-built energy storage system control strategy model simulates supply and demand fluctuations under complex scenarios, dynamically optimizes the charging and discharging plans of energy storage devices, and generates an energy storage resource allocation efficiency optimization scheme, including: The pre-built energy storage system control strategy model is used to simulate the operation state under complex scenarios and determine the operation constraints of energy storage devices in different time periods. Based on the analysis of simulation results, a control strategy framework is constructed to meet the needs of dynamic adjustment. To obtain the optimal response mode of energy storage equipment under supply and demand fluctuations, and to obtain a preliminary scheme for charging and discharging; Based on whether the preliminary scheme exceeds the available capacity limit of the energy storage device, determine whether to adjust the parameters of the control strategy framework to obtain an updated scheme that meets the available capacity limit; Based on the updated scheme and real-time data of power grid operation, short-term changes in supply and demand fluctuations are monitored. If abnormal fluctuations are detected, the parameters of the control strategy framework are adjusted to optimize the updated scheme and obtain an optimized scheme for energy storage resource allocation efficiency. Otherwise, the updated scheme is directly used as the optimized scheme for energy storage resource allocation efficiency.

4. The method as described in claim 1, characterized in that, The method, based on the energy storage resource allocation efficiency optimization scheme, real-time demand for grid supply and demand balance, changes in user electricity consumption behavior, and dynamic grid operation status, employs a game theory-based price adjustment algorithm to generate a real-time electricity price adjustment scheme, including: Based on the energy storage resource allocation efficiency optimization scheme and the real-time demand of power grid supply and demand balance, a dynamic electricity price generation model is constructed. A game theory price adjustment algorithm based on Nash equilibrium is adopted to generate a dynamic electricity price adjustment scheme by combining user electricity consumption behavior, power grid operation dynamics, and power grid supply and demand balance status. Based on the price change trend of the aforementioned dynamic electricity price adjustment scheme, the energy storage resource allocation efficiency optimization scheme is adjusted, and the resource allocation and electricity price linkage scheme is determined. The parameters in the dynamic electricity price generation model are updated by the resource allocation and electricity price linkage scheme. Based on the updated dynamic electricity price generation model, the real-time demand for power grid supply and demand balance is monitored, and a real-time electricity price adjustment scheme is generated.

5. The method as described in claim 4, characterized in that, The aforementioned dynamic electricity price generation model is constructed based on the energy storage resource allocation efficiency optimization scheme and the real-time demand of power grid supply and demand balance. A game-theoretic price adjustment algorithm based on Nash equilibrium is used, combined with user electricity consumption behavior, power grid operation dynamics, and the power grid supply and demand balance state, to generate a dynamic electricity price adjustment scheme, including: Acquire power grid supply and demand data and balance status information, collect electricity load data and energy storage resource status data in the current period, perform data cleaning and formatting on the electricity load data and energy storage resource status data to obtain a standardized supply and demand dataset; The matching degree between the energy storage resource allocation efficiency and the power grid supply and demand data is analyzed based on the supply and demand dataset. When the matching degree is lower than the preset threshold, the allocation efficiency optimization process is triggered. By comparing the operation dynamics in historical data, the adjustment direction of the current energy storage resources is determined. Based on the adjustment direction combined with electricity consumption behavior data and operational dynamic information, the objective function and constraints are determined, and a dynamic electricity price generation model is constructed. Using real-time load data and energy storage dispatch results, the electricity price fluctuation range is set, and the situation where the current load exceeds the average value is analyzed to determine the amount of electricity price adjustment. Based on the dynamic electricity price generation model, a game theory price adjustment algorithm based on Nash equilibrium is used to iteratively calculate the responses of both parties to the electricity price adjustment amount, and then derive the electricity price adjustment plan.

6. The method as described in claim 1, characterized in that, The aforementioned method, based on a real-time electricity price adjustment scheme and combined with the need to protect users' economic interests, generates a user behavior response dataset through a pre-built user behavior response model, including: By using a pre-established electricity price guidance database, we can obtain the specific parameters of the real-time electricity price adjustment plan and historical user feedback data; and classify users according to their economic interests to obtain a preliminary user classification set. Based on historical user feedback data, a logistic regression algorithm is used to predict the distribution of user acceptance of electricity price adjustments, and the prediction results of acceptance are determined. Based on the acceptance prediction results, the likelihood of changes in user behavior is analyzed to obtain the probability distribution of behavior changes. By using the probability distribution of behavioral changes and combining historical data of power grid load, we can simulate the impact trend of user behavior changes on power grid load and determine the key time periods and regional distribution of load changes. For key time periods and regional distributions of load changes, obtain user load change values, construct a correlation mapping between user behavior and load changes, and determine the user behavior response dataset.

7. The method as described in claim 1, characterized in that, The fluctuation rate of the power grid load change trend is calculated by combining the user behavior response dataset with real-time monitoring data of the power grid operation status and the power grid supply and demand balance dynamic assessment algorithm. Based on whether the volatility of the power grid load change trend reaches a preset stability threshold, the final energy storage dispatch scheme is determined, including: Step 1: Use time series analysis to analyze the data in the user behavior response dataset, extract key features of load changes in power grid operation, and determine the periodicity and distribution of abnormal points of load changes. Step II: Based on the periodicity of load changes and the distribution of anomalies, apply the dynamic assessment mechanism of supply and demand balance to calculate the volatility of the power grid load change trend. Step III: Determine whether the volatility of the power grid load change trend is abnormal. If it is not abnormal, the energy storage resource allocation efficiency optimization scheme shall be adopted as the final energy storage scheduling scheme. Otherwise, adjust the energy storage resource allocation efficiency optimization scheme, recalculate the user behavior response dataset, and return to Step I.

8. A control system for an energy storage system based on time-of-use pricing and multimodal optimization, characterized in that, include: The simulation module is used to simulate supply and demand fluctuations in complex scenarios based on a pre-built energy storage system control strategy model, dynamically optimize the charging and discharging plans of energy storage devices, and generate energy storage resource allocation efficiency optimization schemes. The electricity price adjustment module is used to generate real-time electricity price adjustment schemes based on the energy storage resource allocation efficiency optimization scheme, the real-time demand for grid supply and demand balance, changes in user electricity consumption behavior, and the dynamic status of grid operation, using a game theory price adjustment algorithm. The behavior response module is used to generate a user behavior response dataset based on the real-time electricity price adjustment scheme and the user's economic interest protection needs, through a pre-built user behavior response model. The scheme determination module is used to calculate the volatility of the power grid load change trend based on the user behavior response dataset and real-time monitoring data of the power grid operation status, using a dynamic evaluation algorithm for power grid supply and demand balance. Based on whether the volatility of the power grid load change trend reaches a preset stability threshold, the final energy storage scheduling scheme is determined.

9. The system as described in claim 8, characterized in that, The simulation module is specifically used for: The pre-built energy storage system control strategy model is used to simulate the operation state under complex scenarios and determine the operation constraints of energy storage devices in different time periods. Based on the analysis of simulation results, a control strategy framework is constructed to meet the needs of dynamic adjustment. To obtain the optimal response mode of energy storage equipment under supply and demand fluctuations, and to obtain a preliminary scheme for charging and discharging; Based on whether the preliminary scheme exceeds the available capacity limit of the energy storage device, determine whether to adjust the parameters of the control strategy framework to obtain an updated scheme that meets the available capacity limit; Based on the updated scheme and real-time data of power grid operation, short-term changes in supply and demand fluctuations are monitored. If abnormal fluctuations are detected, the parameters of the control strategy framework are adjusted to optimize the updated scheme and obtain an optimized scheme for energy storage resource allocation efficiency. Otherwise, the updated scheme is directly used as the optimized scheme for energy storage resource allocation efficiency.

10. The system as described in claim 8, characterized in that, It also includes a model building module for: Data on power grid operation status, energy storage device operation, and user electricity consumption behavior are acquired, and the data on power grid operation status, energy storage device operation, and user electricity consumption behavior are cleaned, standardized, and formatted to obtain a power grid operation dataset. Data fusion technology is used to fuse the data in the power grid operation dataset, and a deep reinforcement learning algorithm is used to extract the time-series correlation features of multi-dimensional information to generate a power grid operation correlation feature matrix. By utilizing the long short-term memory network's ability to capture long-term dependencies on time-series data, the correlation feature matrix of power grid operation is used to predict the future trends of power grid operation status and user electricity consumption behavior, generate the power grid supply and demand balance demand value and the available capacity range of energy storage devices within the prediction period, and form the power grid operation prediction demand distribution. Based on the distribution of demand predicted by power grid operation and the available capacity range of energy storage devices, a control strategy model for energy storage systems based on deep reinforcement learning is constructed.

11. The system as described in claim 8, characterized in that, The electricity price adjustment module includes: The dynamic adjustment unit is used to construct a dynamic electricity price generation model based on the energy storage resource allocation efficiency optimization scheme and the real-time demand of power grid supply and demand balance. It adopts a game theory price adjustment algorithm based on Nash equilibrium, combined with user electricity consumption behavior, power grid operation dynamics, and power grid supply and demand balance status, to generate a dynamic electricity price adjustment scheme. The linkage unit is used to adjust the energy storage resource allocation efficiency optimization scheme based on the price change trend of the dynamic electricity price adjustment scheme, and determine the resource allocation and electricity price linkage scheme. The monitoring and adjustment unit is used to update the parameters in the dynamic electricity price generation model through the resource allocation and electricity price linkage scheme, monitor the real-time demand for power grid supply and demand balance based on the updated dynamic electricity price generation model, and generate a real-time electricity price adjustment scheme.

12. The system as claimed in claim 11, characterized in that, The dynamic adjustment unit is specifically used for: Acquire power grid supply and demand data and balance status information, collect electricity load data and energy storage resource status data in the current period, perform data cleaning and formatting on the electricity load data and energy storage resource status data to obtain a standardized supply and demand dataset; The matching degree between the energy storage resource allocation efficiency and the power grid supply and demand data is analyzed based on the supply and demand dataset. When the matching degree is lower than the preset threshold, the allocation efficiency optimization process is triggered. By comparing the operation dynamics in historical data, the adjustment direction of the current energy storage resources is determined. Based on the adjustment direction combined with electricity consumption behavior data and operational dynamic information, the objective function and constraints are determined, and a dynamic electricity price generation model is constructed. By using real-time load data and energy storage dispatch results, a range for electricity price fluctuations is set, the situation where the current load exceeds the average value is analyzed, and the amount of electricity price adjustment is determined. Based on the dynamic electricity price generation model, a game theory price adjustment algorithm based on Nash equilibrium is used to iteratively calculate the responses of both parties to the electricity price adjustment amount, and then derive the electricity price adjustment plan.

13. The system as described in claim 8, characterized in that, The behavior response module is specifically used for: By using a pre-established electricity price guidance database, we can obtain the specific parameters of the real-time electricity price adjustment plan and historical user feedback data; and classify users according to their economic interests to obtain a preliminary user classification set. Based on historical user feedback data, a logistic regression algorithm is used to predict the distribution of user acceptance of electricity price adjustments, and the prediction results of acceptance are determined. Based on the acceptance prediction results, the likelihood of changes in user behavior is analyzed to obtain the probability distribution of behavior changes. By using the probability distribution of behavioral changes and combining historical data of power grid load, we can simulate the impact trend of user behavior changes on power grid load and determine the key time periods and regional distribution of load changes. For key time periods and regional distributions of load changes, obtain user load change values, construct a correlation mapping between user behavior and load changes, and determine the user behavior response dataset.

14. The system as described in claim 8, characterized in that, The scheme determination module is specifically used for: Step 1: Use time series analysis to analyze the data in the user behavior response dataset, extract key features of load changes in power grid operation, and determine the periodicity and distribution of abnormal points of load changes. Step II: Based on the periodicity of load changes and the distribution of anomalies, apply the dynamic assessment mechanism of supply and demand balance to calculate the volatility of the power grid load change trend. Step III: Determine whether the volatility of the power grid load change trend is abnormal. If it is not abnormal, the energy storage resource allocation efficiency optimization scheme shall be adopted as the final energy storage scheduling scheme. Otherwise, adjust the energy storage resource allocation efficiency optimization scheme, recalculate the user behavior response dataset, and return to Step I.

15. An electronic device, characterized in that, include: At least one processor and memory; The memory and processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, the energy storage system control method based on time-of-use pricing and multimodal optimization as described in any one of claims 1 to 7 is implemented.

16. A readable storage medium, characterized in that, It contains an execution program, which, when executed, implements the energy storage system control method based on time-of-use pricing and multimodal optimization as described in any one of claims 1 to 7.

Images