Energy storage-photovoltaic-electrolytic aluminum system dynamic optimization method suitable for multiple scenes

Abstract

The invention discloses an energy storage-photovoltaic-electrolytic aluminum system dynamic optimization method suitable for multiple scenes, and relates to the technical field of intelligent power grids, and the method comprises the steps: collecting and preprocessing energy storage-photovoltaic-electrolytic aluminum system data, obtaining photovoltaic, load and electricity price characteristics, constructing an equipment mathematical model, and carrying out the optimization of the equipment mathematical model; the method comprises the steps of predicting a state of charge, battery cycle aging cost, photovoltaic output power, load baseline power and an adjustable flexible interval, constructing a system state prediction model, an economic dispatching optimization model and a self-adaptive control model, outputting a power prediction report, a day-ahead optimal economic dispatching plan and an energy storage power control instruction, and carrying out energy storage control. The energy storage power control instruction is subjected to multi-scene logic judgment and then is transmitted to the energy storage converter to be executed, feedback is obtained, and visualization is carried out through the integrated digital board. Through intelligent prediction and adaptive control, collaborative optimization of economy, reliability and green power consumption efficiency is realized, and the comprehensive performance of the energy storage-photovoltaic-electrolytic aluminum system under variable working conditions is improved.

Description

Technical Field

[0001] The present invention relates to the field of smart grid technology, and in particular to a dynamic optimization method for an energy storage-photovoltaic-electrolytic aluminum system that is adaptable to multiple scenarios. Background Art

[0002] The energy storage-photovoltaic-electrolytic aluminum system is a new power system that integrates energy storage technology, photovoltaic power generation and electrolytic aluminum production. The system builds user-side energy storage projects, uses the photovoltaic power generation absorbed during the day for charging, and discharges during peak load periods of the grid to achieve a steady flow of electricity and peak shaving and valley filling. At the same time, the system actively responds to and promotes the development of the energy storage industry, increases the proportion of green electricity use, and reduces corporate electricity costs. During the electrolytic aluminum production process, the system profits by storing photovoltaic electricity and utilizing the peak-valley electricity price difference, which helps companies achieve green, low-carbon, and high-quality development and serves as a model for the development of the energy storage industry.

[0003] To address the high electricity costs, stringent power supply stability requirements, and difficulty absorbing volatile photovoltaic energy in the high-energy-consuming electrolytic aluminum production model, existing technologies use peak-valley arbitrage strategies based on fixed time periods and simple threshold judgment logic. However, this can lead to rigid control strategies and an inability to adapt in real time to random fluctuations in photovoltaic output and actual load changes. This can lead to insufficient green power absorption, suboptimal overall system economic benefits, and difficulty balancing power supply reliability under variable operating conditions. To address these issues, a dynamic optimization method for energy storage-photovoltaic-electrolytic aluminum systems that is adaptable to multiple scenarios is proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide a dynamic optimization method for an energy storage-photovoltaic-electrolytic aluminum system that is adaptable to multiple scenarios, so as to solve the problems raised in the above background technology.

[0005] To solve the above technical problems, the technical solution adopted by the present invention is: a dynamic optimization method for an energy storage-photovoltaic-electrolytic aluminum system that is adaptable to multiple scenarios, comprising the following steps: Step 1: Collect and preprocess energy storage-photovoltaic-electrolytic aluminum system data to obtain preprocessed energy storage-photovoltaic-electrolytic aluminum system data, and obtain photovoltaic characteristics, load characteristics, and electricity price characteristics from the preprocessed energy storage-photovoltaic-electrolytic aluminum system data, wherein the energy storage-photovoltaic-electrolytic aluminum system data includes energy storage system data, photovoltaic system data, electrolytic aluminum load power data, power grid data, and meteorological data; Step 2: Input the preprocessed energy storage system data and energy storage power control instructions into the energy storage system model, and output the state of charge and battery cycle aging cost at the next moment. Based on the preprocessed meteorological data and the long short-term memory network, the photovoltaic power generation model is trained, and the trained photovoltaic power generation model outputs the photovoltaic output power for the next 24 hours. The load characteristics and factory production plan are input into the electrolytic aluminum load model, and the baseline and flexibility dichotomy method is used to output the load baseline power and adjustable flexibility range. Step 3: The system state prediction model uses the output of the photovoltaic power generation model and the output of the electrolytic aluminum load model to obtain a power forecast report. The economic dispatch optimization model uses mixed integer linear programming based on electricity price characteristics, power forecast reports, and the output of the energy storage system model to output the day-ahead optimal economic dispatch plan. The adaptive control model uses deep reinforcement learning technology based on preprocessed photovoltaic system data, energy storage system data, electrolytic aluminum load power data, and the day-ahead optimal economic dispatch plan to output energy storage power control instructions. Step 4: Combine the grid data and meteorological data, transmit the energy storage power control command to the energy storage converter for execution after multi-scenario logic judgment, and obtain the execution effect of the energy storage power control command; Step 5: Visualize the energy storage charge state, photovoltaic power, and photovoltaic and load performance curve data through the integrated digital dashboard, and feed back the execution effect of the energy storage power control command to the adaptive control model.

[0006] A further improvement of the technical solution of the present invention is that: the energy storage-photovoltaic-electrolytic aluminum system data includes energy storage system data, photovoltaic system data, electrolytic aluminum load power data, power grid data and meteorological data; the energy storage-photovoltaic-electrolytic aluminum system data is collected and preprocessed to obtain preprocessed energy storage-photovoltaic-electrolytic aluminum system data; and the process of obtaining photovoltaic characteristics, load characteristics and electricity price characteristics from the preprocessed energy storage-photovoltaic-electrolytic aluminum system data includes: Within the project plant's integrated energy storage converter and booster, the integrated energy storage converter and battery management system collect the system's real-time charge and discharge power and state of charge as energy storage system data. The energy management system regularly polls the energy storage converter and battery management system via an internal communication network using standard protocols to obtain this data. By installing a multi-function electric energy meter at the grid-connected junction point of the photovoltaic area within the project plant, the photovoltaic power time series of the photovoltaic system is measured in real time as photovoltaic system data. The energy management system reads the data of the multi-function electric energy meter through the communication interface to grasp the real-time photovoltaic power generation situation; A high-precision power quality monitoring device is installed on the feeder circuit of the power distribution room in the project plant area that supplies power to the electrolytic aluminum load. The power quality monitoring device monitors the real-time active power and current flowing to the electrolytic aluminum load as electrolytic aluminum load power data. The energy management system obtains the power consumption of the core load by reading the data from the power quality monitoring device; The measurement, control and protection devices at the project plant's grid connection point are used to collect the grid exchange power, frequency and voltage. The time-of-use electricity price list is statically configured as basic information in the energy management system, with the grid exchange power, frequency and voltage and the time-of-use electricity price list as grid data. A small automated weather station deployed in the photovoltaic area collects real-time total solar radiation intensity, ambient temperature, relative humidity and wind speed at the project site as meteorological data; The collected energy storage system data, photovoltaic system data, electrolytic aluminum load power data, power grid data, and meteorological data are timestamped and transmitted to the energy management system's historical database for centralized storage. The energy storage-photovoltaic-electrolytic aluminum system data is cleaned. The cleaning process includes identifying outlier data points caused by sensor jumps and communication interruptions using the interquartile range method, filling these outlier data points with valid data points at adjacent moments through linear interpolation, and time-aligning the cleaned energy storage-photovoltaic-electrolytic aluminum system data. Maximum and minimum normalization is then performed to obtain preprocessed energy storage system data, photovoltaic system data, electrolytic aluminum load power data, power grid data, and meteorological data. Photovoltaic characteristics include the photovoltaic power change rate and short-term volatility characteristics. The photovoltaic power change rate is calculated based on the preprocessed photovoltaic system data and the difference between the power value at the current moment and the power value at the previous moment. The short-term volatility characteristic is obtained by calculating the standard deviation of the photovoltaic system data within a preset time window. The load characteristics include the load baseline level and the high-frequency disturbance component characteristics. The real-time active power is smoothed using a moving average filter. The output of the moving average filter is regarded as the load baseline level. The difference sequence obtained by subtracting the load baseline level from the real-time active power is used to characterize the high-frequency disturbance component characteristics. The electricity price characteristics include the peak and valley period distribution and the peak and valley price difference characteristics in the next 24 hours. The program parses the time-of-use electricity price table, identifies and outputs the start and end time points of the peak, valley and flat periods in the next 24 hours, and forms the peak and valley period distribution in the next 24 hours. Based on the peak and valley period distribution in the next 24 hours, the difference between the peak period electricity price and the valley period electricity price is extracted and calculated from the configured time-of-use electricity price table to obtain the peak and valley price difference characteristics.

[0007] A further improvement of the technical solution of the present invention is that the process of inputting the pre-processed energy storage system data and the energy storage power control instruction into the energy storage system model and outputting the state of charge and battery cycle aging cost at the next moment includes: The energy storage system model receives energy storage power control instructions from the adaptive control model If the energy storage power control instruction is negative, it is regarded as charging power If the energy storage power control instruction is positive, it is regarded as the discharge power , the rated capacity of the energy storage system is preset , charging efficiency and discharge efficiency , calculate the state of charge at the next moment , the calculation process is as follows: ; in, is the state of charge at the previous moment, is the time step of the calculation; Total investment cost based on pre-set project site and the number of cycles in the battery's full life cycle , calculate the equivalent cycle cost of completing a full charge and discharge , according to the current charge and discharge energy , calculate the equivalent number of cycles of charge and discharge energy , multiply the equivalent number of cycles by the equivalent cycle cost to obtain the battery cycle aging cost in the current time step , the calculation process is as follows: ; ; in, is the total energy throughput of a full charge and discharge.

[0008] A further improvement of the technical solution of the present invention is that the process of training a photovoltaic power generation model based on preprocessed meteorological data and a long short-term memory network to output photovoltaic output power for the next 24 hours includes: A photovoltaic power generation model based on a long short-term memory network is trained and fixed in the energy management system. It is executed at a fixed frequency. Meteorological data is used as the input features of the photovoltaic power generation model and is divided into a training set and a validation set. The actual total output power of the photovoltaic power generation system at the same time is used as the output label. The long short-term memory network includes an input layer, a hidden layer, and an output layer. The number of nodes in the input layer corresponds to the dimension of the meteorological data; The hidden layer consists of a long short-term memory layer containing a gating structure. The gating structure selectively memorizes and forgets historical information received from the input layer, thereby capturing the nonlinear mapping relationship between meteorological data and photovoltaic output power. The number of nodes in the output layer corresponds to the future time step that needs to be predicted, and the photovoltaic output power for the next 24 hours is output; The training set is input into the trained long short-term memory network, and the mean square error between the output photovoltaic output power for the next 24 hours and the actual output label is calculated. Adaptive estimation is used to adjust the weight parameters within the long short-term memory network according to the back propagation of the mean square error. The training set calculation process is iterated until the mean square error converges, and the generalization ability of the photovoltaic power generation model is tested using the validation set.

[0009] A further improvement of the technical solution of the present invention is that the load characteristics and the factory production plan are input into the electrolytic aluminum load model, and the process of outputting the load baseline power and the adjustable flexibility range using the baseline and flexibility dichotomy method includes: The load baseline level in the load characteristic is taken as the load baseline power; The distribution range of high-frequency disturbance component characteristics during normal production is statistically analyzed to obtain the initial adjustment range driven by historical data. The initial adjustment range is compared with the rigid safety boundary set according to the factory production plan and the electrolytic cell process requirements. The intersection of the initial adjustment range and the rigid safety boundary is taken as the output adjustable flexible range.

[0010] A further improvement of the technical solution of the present invention is that the process of obtaining a power forecast report by utilizing the output of the photovoltaic power generation model and the output of the electrolytic aluminum load model includes: The system state prediction model uses the photovoltaic output power time series for the next 24 hours output by the photovoltaic power generation model and the load baseline power output by the electrolytic aluminum load model, and extrapolates the most recent load baseline power to the next 24 hours to form the load baseline power time series for the next 24 hours. The load baseline power time series and the photovoltaic output power time series are calculated in parallel. Subtract the photovoltaic output power time series , obtain the net load forecast sequence of the energy storage-photovoltaic-electrolytic aluminum system in the next 24 hours, and integrate the photovoltaic output power time series, load baseline power time series and net load forecast sequence into a power forecast report.

[0011] A further improvement of the technical solution of the present invention is that: based on the economic dispatch optimization model, electricity price characteristics, power forecast reports and the output of the energy storage system model, mixed integer linear programming is used to output the optimal economic dispatch plan for the day ahead, including the following process: Economic dispatch optimization model based on grid exchange power , real-time time-of-use electricity prices provided by the time-of-use electricity price table and battery cycle aging costs , using mixed integer linear programming, with the lowest daily operating cost of the energy storage-photovoltaic-electrolytic aluminum system as the objective function Optimize the solution, where For the timing, it means purchasing electricity. When it is negative, it indicates that electricity is sold and fed back to the grid, and T represents the total number of time steps in the optimization cycle; Charging power , discharge power The binary variables that make charging and discharging occur at different times are used as decision variables to establish power balance constraints and energy storage system operation constraints; Power balance constraints , used to balance the power generation and consumption of the energy storage system; Energy storage system operation constraints include state of charge update constraints, state of charge boundary constraints and charge and discharge power constraints. , which is used to update the state of charge of the energy storage system according to the charge and discharge power at any time t, and the state of charge boundary constraint , used to preset the state of charge of the energy storage system between the minimum and maximum values, and the charge and discharge power constraints , , which is used to limit the charging and discharging power of the energy storage system to a preset maximum value, and charging and discharging cannot be carried out at the same time, where and is the minimum state of charge and the maximum state of charge, and are the maximum values ​​of charging power and discharging power respectively; The objective function, decision variables, and energy storage system operating constraints are integrated, and the decision variable sequence that minimizes the objective function value is calculated through mixed integer linear programming. The decision variable sequence that minimizes the objective function value is used as the optimal charge and discharge power of the energy storage system at each time step in the next 24 hours. The optimal charge and discharge power at each time step in the next 24 hours is formatted as the day-ahead optimal economic dispatch plan.

[0012] A further improvement to the technical solution of the present invention is that, through an adaptive control model, based on pre-processed photovoltaic system data, energy storage system data, electrolytic aluminum load power data, and the day-ahead optimal economic dispatch plan, deep reinforcement learning technology is used to output energy storage power control instructions, including the following process: Define the state, action, and reward required for the adaptive control model to make energy storage power control command decisions. The state includes the real-time photovoltaic power of the photovoltaic system, the real-time electrolytic aluminum load power, and the power deviation from the optimal economic dispatch plan at the current moment. The action is the energy storage power control command for the next time step whose value range is constrained by the rated power of the energy storage converter. The reward is a negative value related to real-time operating costs, where represents the actual cost of purchasing electricity from the grid at time t; The digital twin system consists of an energy storage system model, a photovoltaic power generation model, and an electrolytic aluminum load model. The intelligent agent is trained in the digital twin system. The intelligent agent outputs energy storage power control instructions, and the digital twin system calculates the state and reward at the next moment and feeds it back to the intelligent agent. The agent is trained using a deep deterministic policy gradient algorithm. The agent simulates interactions within the digital twin system. In each simulated interaction, the agent selects an action based on its current state. The digital twin system then provides feedback on the reward and new state. The agent then adjusts the weight parameters of its internal neural network based on the reward signal. The agent training process is iterated repeatedly until the agent's strategy converges and the maximum cumulative reward is continuously achieved. The mature intelligent agent strategy network trained offline is deployed in the energy management system. At each time step of actual operation, the energy management system inputs the real-time collected state data into the intelligent agent strategy network. The intelligent agent strategy network performs a forward propagation calculation and outputs the optimal energy storage power control instruction under the current state.

[0013] A further improvement of the technical solution of the present invention is that the multiple scenarios include normal scenarios, extreme weather scenarios, and grid failure scenarios. The process of combining grid data and weather data to transmit the energy storage power control command to the energy storage converter for execution after multi-scenario logic judgment includes: When the grid data is within the normal operating range and the energy management system has not received any external extreme weather warning signals, the current scenario is determined to be a normal scenario, and the energy storage power control instructions output by the adaptive control model are directly transmitted to the energy storage converter for execution; When the energy management system receives an external warning based on meteorological data, it determines that the current scenario is an extreme weather scenario, temporarily suspends the energy storage power control instructions output by the adaptive control model, executes the emergency plan, generates safety-oriented energy storage power control instructions, and transmits the safety-oriented energy storage power control instructions to the energy storage converter for execution; When the measurement, control and protection device at the grid connection point detects an anomaly in the grid data, it determines that the current scenario is a grid failure scenario, triggers the highest priority protection and control logic, and issues an instruction to disconnect the circuit breaker connected to the public grid, causing the plant microgrid to enter island operation mode. The energy storage power control instruction output by the adaptive control model is terminated, and the power supply control logic for island operation is switched to maintain the voltage and frequency stability of the grid in the island, giving priority to ensuring continuous power supply to the core load of electrolytic aluminum. The island operation controller generates a balanced energy storage power control instruction based on the real-time power balance requirement, and transmits the balanced energy storage power control instruction to the energy storage converter.

[0014] A further improvement of the technical solution of the present invention is that the process of visualizing the energy storage charge state, photovoltaic power, and photovoltaic and load performance curve data through an integrated digital dashboard and feeding back the execution effect of the energy storage power control instruction to the adaptive control model includes: The front-end application of the energy management system requests and aggregates energy storage state of charge, photovoltaic power, and photovoltaic and load performance curve data from historical and real-time databases, and then structures and visualizes them on the human-computer interaction interface through dynamic charts, dashboards, and status indicators; After each energy storage power control instruction is executed, the energy management system records the experience data unit including the state before execution, the executed action, the actual reward and the new state after execution, and stores the experience data unit in the experience replay database. Data is regularly extracted from the experience replay database to perform offline retraining and optimization of the policy network of the adaptive control model, and the updated policy network is redeployed back to the energy management system.

[0015] Due to the adoption of the above technical solution, the present invention has the following technical advancements compared to the prior art: 1. The present invention provides a dynamic optimization method for an energy storage-photovoltaic-electrolytic aluminum system that is adaptable to multiple scenarios. By constructing a prediction model and an adaptive control model, it achieves accurate prediction of photovoltaic output and real-time dynamic optimization of energy storage charging and discharging strategies. It can maximize the use of photovoltaic green electricity for peak-valley arbitrage based on actual lighting conditions, and comprehensively considers the battery cycle aging cost, significantly improving the overall economic benefits of the entire system.

[0016] 2. The present invention provides a dynamic optimization method for an energy storage-photovoltaic-electrolytic aluminum system that is adaptable to multiple scenarios. By introducing multi-scenario judgment logic, the system can adaptively switch control targets between different modes of conventional economic operation, extreme weather, and power grid failures, achieving a dynamic balance from a single economic goal to a multi-objective balance that takes reliability into account, giving priority to ensuring the power supply stability and safety of the core load of electrolytic aluminum, and effectively enhancing the system's operational reliability under complex working conditions.

[0017] 3. The present invention provides a dynamic optimization method for energy storage-photovoltaic-electrolytic aluminum systems that is adaptable to multiple scenarios. By establishing a complete intelligent workflow from data acquisition, feature extraction to closed-loop feedback, it achieves coordinated and refined management of photovoltaic, energy storage and load units, solves the problem of insufficient green energy consumption caused by the inability of fixed control strategies to match real-time operating conditions, and comprehensively improves energy utilization efficiency and the level of system automation operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0019] Figure 1 A flow chart of a dynamic optimization method for an energy storage-photovoltaic-electrolytic aluminum system adapted to multiple scenarios provided by the present invention. DETAILED DESCRIPTION

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0021] Example 1, as Figure 1 As shown, the present invention provides a dynamic optimization method for an energy storage-photovoltaic-electrolytic aluminum system that is adaptable to multiple scenarios, comprising the following steps: Step 1: Collect and preprocess the energy storage-photovoltaic-electrolytic aluminum system data to obtain preprocessed energy storage-photovoltaic-electrolytic aluminum system data, and obtain photovoltaic characteristics, load characteristics, and electricity price characteristics from the preprocessed energy storage-photovoltaic-electrolytic aluminum system data. The energy storage-photovoltaic-electrolytic aluminum system data includes energy storage system data, photovoltaic system data, electrolytic aluminum load power data, power grid data, and meteorological data.

[0022] In some embodiments, an integrated energy storage converter and battery management system are used inside the energy storage converter and booster in the project plant to collect the real-time charging and discharging power and state of charge of the energy storage system as energy storage system data. The energy management system polls the energy storage converter and battery management system regularly through the internal communication network in accordance with standard protocols to obtain the energy storage system data.

[0023] In some embodiments, a multi-function electric energy meter is installed at the grid-connected junction of the photovoltaic area within the project plant to measure the photovoltaic power time series of the photovoltaic system in real time as photovoltaic system data. The energy management system reads the data of the multi-function electric energy meter through the communication interface to grasp the real-time photovoltaic power generation situation.

[0024] In some embodiments, a high-precision power quality monitoring device is installed on the feeder loop of the power distribution room in the project plant area that supplies power to the electrolytic aluminum load. The power quality monitoring device monitors the real-time active power and current flowing to the electrolytic aluminum load as the electrolytic aluminum load power data. The energy management system obtains the power consumption of the core load by reading the data from the power quality monitoring device.

[0025] In some embodiments, the measurement, control and protection device at the project plant grid connection point is used to collect the grid exchange power, frequency and voltage, and the time-of-use electricity price table is statically configured in the energy management system as basic information, and the grid exchange power, frequency and voltage and the time-of-use electricity price table are used as grid data.

[0026] In some embodiments, a small automated weather station is deployed in the photovoltaic area to collect real-time total solar radiation intensity, ambient temperature, relative humidity and wind speed at the project site as meteorological data.

[0027] In some embodiments, the collected energy storage system data, photovoltaic system data, electrolytic aluminum load power data, power grid data and meteorological data are attached with a unified timestamp and transmitted to the historical database of the energy management system for centralized storage. The energy storage-photovoltaic-electrolytic aluminum system data is cleaned. The cleaning process includes identifying outlier data points caused by sensor jumps and communication interruptions through the interquartile range method, filling the outlier data points with linear interpolation of valid data points at adjacent moments, aligning the cleaned energy storage-photovoltaic-electrolytic aluminum system data in time scales, and performing maximum and minimum normalization processing to obtain preprocessed energy storage system data, photovoltaic system data, electrolytic aluminum load power data, power grid data and meteorological data.

[0028] In some embodiments, the photovoltaic characteristics include a photovoltaic power change rate and a short-term volatility characteristic, wherein the photovoltaic power change rate is obtained by calculating the difference between the power value at the current moment and the power value at the previous moment based on the preprocessed photovoltaic system data, and the short-term volatility characteristic is obtained by calculating the standard deviation of the photovoltaic system data within a preset time window within a preset time window.

[0029] In some embodiments, the load characteristics include a load baseline level and high-frequency disturbance component characteristics. The real-time active power is smoothed by a moving average filter, the output of the moving average filter is regarded as the load baseline level, the real-time active power is subtracted from the load baseline level, and the resulting difference sequence is used to characterize the high-frequency disturbance component characteristics.

[0030] In some embodiments, the electricity price characteristics include the peak and valley period distribution and the peak and valley price difference characteristics within the next 24 hours. The program parses the time-of-use electricity price table, identifies and outputs the start and end time points of the peak, valley and flat periods within the next 24 hours, and forms the peak and valley period distribution within the next 24 hours. Based on the peak and valley period distribution within the next 24 hours, the difference between the numerical values ​​of the peak period electricity price and the valley period electricity price is extracted and calculated from the configured time-of-use electricity price table to obtain the peak and valley price difference characteristics.

[0031] In step 2, the preprocessed energy storage system data and energy storage power control instructions are input into the energy storage system model to output the state of charge and battery cycle aging cost at the next moment; based on the preprocessed meteorological data and the long short-term memory network, the photovoltaic power generation model is trained, and the photovoltaic output power for the next 24 hours is output through the trained photovoltaic power generation model; the load characteristics and factory production plan are input into the electrolytic aluminum load model, and the baseline and flexibility dichotomy method is used to output the load baseline power and adjustable flexibility range.

[0032] In some embodiments, the energy storage system model receives energy storage power control instructions from the adaptive control model If the energy storage power control instruction is negative, it is regarded as charging power If the energy storage power control instruction is positive, it is regarded as the discharge power , the rated capacity of the energy storage system is preset , charging efficiency and discharge efficiency , calculate the state of charge at the next moment , the specific calculation formula is as follows:

[0033] in, is the state of charge at the previous moment, is the time step of the calculation.

[0034] In some embodiments, based on the total investment cost of the project site and the number of cycles in the battery's full life cycle , calculate the equivalent cycle cost of completing a full charge and discharge , according to the current charge and discharge energy , calculate the equivalent number of cycles of charge and discharge energy , multiply the equivalent number of cycles by the equivalent cycle cost to obtain the battery cycle aging cost in the current time step , the specific calculation formula is as follows:

[0035]

[0036] in, is the total energy throughput of a full charge and discharge.

[0037] In some embodiments, a photovoltaic power generation model based on a long short-term memory network is trained and fixed in an energy management system, and is executed in a rolling manner at a fixed frequency. Meteorological data is used as input features of the photovoltaic power generation model and is divided into a training set and a validation set. The actual total output power of the photovoltaic power generation system corresponding to the same moment is used as the output label. The long short-term memory network includes an input layer, a hidden layer, and an output layer.

[0038] In some embodiments, the number of nodes in the input layer corresponds to the dimension of the meteorological data.

[0039] In some embodiments, the hidden layer is composed of a long short-term memory layer including a gating structure, which selectively memorizes and forgets historical information received from the input layer, thereby capturing the nonlinear mapping relationship between meteorological data and photovoltaic output power.

[0040] In some embodiments, the number of nodes in the output layer corresponds to the future time step that needs to be predicted, and the photovoltaic output power for the next 24 hours is output.

[0041] In some embodiments, the training set is input into the trained long short-term memory network, and the mean square error between the output photovoltaic output power for the next 24 hours and the actual output label is calculated. Adaptive estimation is used to adjust the weight parameters inside the long short-term memory network according to the back propagation of the mean square error. The training set calculation process is iterated until the mean square error converges, and the validation set is used to test the generalization ability of the photovoltaic power generation model.

[0042] In some embodiments, the load baseline level in the load characteristic is used as the load baseline power.

[0043] In some embodiments, the distribution range of high-frequency disturbance component characteristics during normal production is statistically analyzed to obtain an initial adjustment range driven by historical data. The initial adjustment range is compared with a rigid safety boundary set according to the factory production plan and the electrolytic cell process requirements, and the intersection of the initial adjustment range and the rigid safety boundary is taken as the output adjustable flexible range.

[0044] Step 3: Through the system state prediction model, the output of the photovoltaic power generation model and the output of the electrolytic aluminum load model are used to obtain a power forecast report; through the economic dispatch optimization model, based on the electricity price characteristics, the power forecast report and the output of the energy storage system model, mixed integer linear programming is used to output the optimal economic dispatch plan on the day before; through the adaptive control model, based on the pre-processed photovoltaic system data, energy storage system data, electrolytic aluminum load power data and the optimal economic dispatch plan on the day before, deep reinforcement learning technology is used to output the energy storage power control instruction.

[0045] In some embodiments, the system state prediction model calls the photovoltaic output power time series of the next 24 hours output by the photovoltaic power generation model and calls the load baseline power output by the electrolytic aluminum load model, and extrapolates the most recent load baseline power to the next 24 hours to form the load baseline power time series of the next 24 hours.

[0046] In some embodiments, the load baseline power time series and the photovoltaic output power time series are calculated in parallel. Subtract the photovoltaic output power time series , obtain the net load forecast sequence of the energy storage-photovoltaic-electrolytic aluminum system in the next 24 hours, and integrate the photovoltaic output power time series, load baseline power time series and net load forecast sequence into a power forecast report.

[0047] In some embodiments, the economic dispatch optimization model is based on the grid exchange power , real-time time-of-use electricity prices provided by the time-of-use electricity price table and battery cycle aging costs , using mixed integer linear programming, with the lowest daily operating cost of the energy storage-photovoltaic-electrolytic aluminum system as the objective function Optimize the solution, where For the timing, it means purchasing electricity. When it is negative, it indicates that electricity is sold and fed back to the grid, and T represents the total number of time steps in the optimization period.

[0048] In some embodiments, the charging power , discharge power The binary variables that make charging and discharging occur at different times are used as decision variables to establish power balance constraints and energy storage system operation constraints.

[0049] In some embodiments, the power balance constraint , used to balance the power generation and consumption of the energy storage system.

[0050] In some embodiments, the energy storage system operation constraints include state of charge update constraints, state of charge boundary constraints and charge and discharge power constraints. , which is used to update the state of charge of the energy storage system according to the charge and discharge power at any time t, and the state of charge boundary constraint , used to preset the state of charge of the energy storage system between the minimum and maximum values, and the charge and discharge power constraints , , which is used to limit the charging and discharging power of the energy storage system to a preset maximum value, and charging and discharging cannot be carried out at the same time, where and is the minimum state of charge and the maximum state of charge, and are the maximum values ​​of charging power and discharging power respectively.

[0051] In some embodiments, the objective function, decision variables, and operating constraints of the energy storage system are integrated, and a decision variable sequence that minimizes the objective function value is calculated through mixed integer linear programming. The decision variable sequence that minimizes the objective function value is used as the optimal charge and discharge power of the energy storage system at each time step in the next 24 hours, and the optimal charge and discharge power at each time step in the next 24 hours is formatted as a day-ahead optimal economic dispatch plan.

[0052] In some embodiments, the state, action, and reward required for the adaptive control model to make energy storage power control instruction decisions are defined. The state includes the real-time photovoltaic power of the photovoltaic system at the current moment, the real-time electrolytic aluminum load power, and the power deviation from the optimal economic dispatch plan of the day before. The action is the energy storage power control instruction for the next time step whose value range is constrained by the rated power of the energy storage converter. The reward is a negative value related to real-time operating costs, where Represents the actual cost of purchasing electricity from the grid at time t.

[0053] In some embodiments, a digital twin system is composed of an energy storage system model, a photovoltaic power generation model, and an electrolytic aluminum load model. An intelligent agent is trained in the digital twin system, and the intelligent agent outputs energy storage power control instructions. The digital twin system calculates the state and reward at the next moment and feeds back to the intelligent agent.

[0054] In some embodiments, a deep deterministic policy gradient algorithm is used to train the intelligent agent, and the intelligent agent performs simulated interactions in the digital twin system. In each simulated interaction, the intelligent agent selects an action based on the current state, and the digital twin system feeds back rewards and new states. The intelligent agent adjusts the weight parameters of its internal neural network based on the reward signal, and repeatedly iterates the intelligent agent training process until the intelligent agent's strategy converges and continuously obtains the maximum cumulative reward.

[0055] In some embodiments, an intelligent agent strategy network that has been maturely trained offline is deployed in an energy management system. At each time step of actual operation, the energy management system inputs the real-time collected state data into the intelligent agent strategy network. The intelligent agent strategy network performs a forward propagation calculation and outputs the optimal energy storage power control instruction under the current state.

[0056] Step 4: Combine grid data with meteorological data. Multiple scenarios include normal scenarios, extreme weather scenarios, and grid failure scenarios. After multi-scenario logic judgment, the energy storage power control command is transmitted to the energy storage converter for execution to obtain the execution effect of the energy storage power control command.

[0057] In some embodiments, when the grid data is within the normal operating range and the energy management system does not receive an external extreme weather warning signal, the current scenario is determined to be a normal scenario, and the energy storage power control instruction output by the adaptive control model is directly transmitted to the energy storage converter for execution.

[0058] In some embodiments, when the energy management system receives an external warning based on meteorological data, it determines that the current scenario is an extreme meteorological scenario, temporarily suspends the energy storage power control instructions output by the adaptive control model, executes the emergency plan, generates safety-oriented energy storage power control instructions, and transmits the safety-oriented energy storage power control instructions to the energy storage converter for execution.

[0059] In some embodiments, when the measurement, control and protection device at the grid connection point detects an abnormality in the grid data, it determines that the current scenario is a grid failure scenario, triggers the highest priority protection and control logic, issues an instruction to disconnect the circuit breaker connected to the public grid, and puts the plant microgrid into island operation mode. The energy storage power control instruction output by the adaptive control model is terminated, and the power supply control logic for island operation is switched to maintain the voltage and frequency stability of the grid in the island, giving priority to ensuring continuous power supply to the core load of electrolytic aluminum. A balanced energy storage power control instruction is generated by the island operation controller according to the real-time power balance requirement, and the balanced energy storage power control instruction is transmitted to the energy storage converter.

[0060] In some embodiments, when the measurement, control and protection device at the grid connection point detects an abnormality in the grid data, it is determined to be a grid fault scenario, triggering the highest priority protection and control logic, issuing an instruction to disconnect the circuit breaker connected to the public grid, causing the plant microgrid to enter island operation mode, terminating the energy storage power control instruction output by the adaptive control model, and switching to the island operation power supply control logic that maintains the voltage and frequency stability of the grid in the island, giving priority to ensuring continuous power supply to the core load of electrolytic aluminum, generating an energy storage power control instruction generated by the island operation controller based on real-time power balance requirements, and transmitting the energy storage power control instruction to the energy storage converter.

[0061] Step 5: Visualize the energy storage charge state, photovoltaic power, and photovoltaic and load performance curve data through the integrated digital dashboard, and feed back the execution effect of the energy storage power control command to the adaptive control model.

[0062] In some embodiments, the front-end application of the energy management system requests and aggregates energy storage charge status, photovoltaic power, and photovoltaic and load performance curve data from historical and real-time databases, and then structures them for visualization on the human-computer interaction interface through dynamic charts, dashboards, and status indicators.

[0063] In some embodiments, after each energy storage power control instruction is executed, the energy management system records an experience data unit including the state before execution, the action executed, the actual reward, and the new state after execution, and stores the experience data unit in an experience replay database. Data is regularly extracted from the experience replay database to perform offline retraining and optimization on the policy network of the adaptive control model, and the updated policy network is redeployed back to the energy management system.

[0064] The above are only specific embodiments of the present application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A dynamic optimization method for energy storage-photovoltaic-aluminum electrolysis system adaptable to multiple scenarios, characterized by: The following steps are involved: collecting and preprocessing energy storage-photovoltaic-electrolytic aluminum system data to obtain preprocessed energy storage-photovoltaic-electrolytic aluminum system data, and obtaining photovoltaic characteristics, load characteristics, and electricity price characteristics from the preprocessed energy storage-photovoltaic-electrolytic aluminum system data, wherein the energy storage-photovoltaic-electrolytic aluminum system data includes energy storage system data, photovoltaic system data, electrolytic aluminum load power data, power grid data, and meteorological data; The pre-processed energy storage system data and energy storage power control instructions are input into the energy storage system model to output the state of charge and battery cycle aging cost at the next moment; based on the pre-processed meteorological data and the long short-term memory network, a photovoltaic power generation model is trained, and the photovoltaic output power for the next 24 hours is output through the trained photovoltaic power generation model; the load characteristics and factory production plan are input into the electrolytic aluminum load model, and the baseline and flexibility dichotomy method is used to output the load baseline power and the adjustable flexibility range; A power forecast report is obtained by using the output of the photovoltaic power generation model and the output of the electrolytic aluminum load model through a system state prediction model; a day-ahead optimal economic dispatch plan is output by using a mixed integer linear programming based on the electricity price characteristics, the power forecast report, and the output of the energy storage system model through an economic dispatch optimization model; and an energy storage power control instruction is output by using deep reinforcement learning technology based on the preprocessed photovoltaic system data, energy storage system data, electrolytic aluminum load power data, and the day-ahead optimal economic dispatch plan through an adaptive control model; Combining the power grid data with the meteorological data, transmitting the energy storage power control instruction to the energy storage converter for execution after multi-scenario logic judgment, and obtaining the execution effect of the energy storage power control instruction; The energy storage charge state, the photovoltaic power, and the photovoltaic and load performance curve data are visualized through an integrated digital dashboard, and the execution effect of the energy storage power control instruction is fed back to the adaptive control model.

2. The method for dynamic optimization of a multi-scenario energy storage-photovoltaic-aluminum electrolysis system according to claim 1, characterized in that: The energy storage-photovoltaic-electrolytic aluminum system data includes energy storage system data, photovoltaic system data, electrolytic aluminum load power data, power grid data, and meteorological data. The process of collecting and preprocessing the energy storage-photovoltaic-electrolytic aluminum system data to obtain the preprocessed energy storage-photovoltaic-electrolytic aluminum system data and obtaining photovoltaic characteristics, load characteristics, and electricity price characteristics from the preprocessed energy storage-photovoltaic-electrolytic aluminum system data includes: Inside the energy storage converter and booster integrated unit at the project site, the integrated energy storage converter and battery management system are used to collect the real-time charge and discharge power and state of charge of the energy storage system as the energy storage system data; By installing a multifunctional electric energy meter at the grid-connected junction point of the photovoltaic area within the project plant, the photovoltaic power time series of the photovoltaic system is measured in real time as the photovoltaic system data; A power quality monitoring device is installed on the feeder circuit of the power distribution room in the project plant area that supplies power to the electrolytic aluminum load. The power quality monitoring device monitors the real-time active power and current flowing to the electrolytic aluminum load as the power data of the electrolytic aluminum load; Utilize the measurement, control and protection device at the project plant's grid connection point to collect the grid exchange power, frequency and voltage, statically configure a time-of-use electricity price list as basic information in a preset energy management system, and use the grid exchange power, frequency and voltage and the time-of-use electricity price list as grid data; By deploying a small automated weather station in the photovoltaic area, the real-time total solar radiation intensity, ambient temperature, relative humidity and wind speed at the project site are collected as the meteorological data; The collected energy storage system data, photovoltaic system data, electrolytic aluminum load power data, power grid data and meteorological data are affixed with a unified timestamp and transmitted to the historical database of the energy management system for centralized storage, and subjected to maximum and minimum normalization processing to obtain the pre-processed energy storage system data, photovoltaic system data, electrolytic aluminum load power data, power grid data and meteorological data; The photovoltaic characteristics include a photovoltaic power change rate and a short-term volatility characteristic, wherein the photovoltaic power change rate is obtained by calculating the difference between the power value at the current moment and the power value at the previous moment based on the preprocessed photovoltaic system data, and the short-term volatility characteristic is obtained by calculating the standard deviation of the photovoltaic system data within a preset time window; The load characteristics include a load baseline level and a high-frequency disturbance component characteristic. The real-time active power is smoothed using a moving average filter. The output of the moving average filter is regarded as the load baseline level. The load baseline level is subtracted from the real-time active power, and the resulting difference sequence is used to characterize the high-frequency disturbance component characteristic. The electricity price characteristics include the peak-valley time period distribution and the peak-valley price difference characteristics within the next 24 hours. The time-of-use electricity price table is parsed by a program to identify and output the start and end time points of the peak, valley and flat time periods within the next 24 hours to form the peak-valley time period distribution within the next 24 hours. Based on the peak-valley time period distribution within the next 24 hours, the difference between the numerical values ​​of the peak time period electricity price and the valley time period electricity price is extracted and calculated from the configured time-of-use electricity price table to obtain the peak-valley price difference characteristics.

3. The method for dynamic optimization of a multi-scenario energy storage-photovoltaic-aluminum electrolysis system according to claim 1, characterized in that: The process of inputting the pre-processed energy storage system data and the energy storage power control instruction into the energy storage system model and outputting the state of charge and battery cycle aging cost at the next moment includes: The energy storage system model receives the energy storage power control instruction from the adaptive control model, and if the energy storage power control instruction is a negative value, it is regarded as charging power; if the energy storage power control instruction is a positive value, it is regarded as discharging power. The energy storage system model calculates the state of charge at the next moment by applying the preset rated capacity, charging efficiency, and discharging efficiency of the energy storage system; Based on the preset total investment cost of the project plant and the number of cycles in the battery's full life cycle, the equivalent cycle cost of completing a full charge and discharge is calculated. According to the charge and discharge energy at the current moment, the equivalent number of cycles of the charge and discharge energy is calculated. The equivalent number of cycles is multiplied by the equivalent cycle cost to obtain the battery cycle aging cost in the current time step.

4. The method for dynamic optimization of a multi-scenario energy storage-photovoltaic-aluminum electrolysis system according to claim 1, characterized in that: The process of training the photovoltaic power generation model based on the pre-processed meteorological data and the long short-term memory network to output the photovoltaic output power for the next 24 hours includes: A photovoltaic power generation model based on a long short-term memory network is trained and fixed in the energy management system, with the model being executed in a rolling manner at a fixed frequency. The meteorological data is used as input features of the photovoltaic power generation model and divided into a training set and a validation set. The actual total output power of the photovoltaic power generation system corresponding to the same time is used as the output label. The long short-term memory network includes an input layer, a hidden layer, and an output layer. The number of nodes in the input layer corresponds to the dimension of the meteorological data; The hidden layer is composed of a long short-term memory layer including a gating structure, wherein the gating structure selectively memorizes and forgets historical information received from the input layer, thereby capturing the nonlinear mapping relationship between the meteorological data and the photovoltaic output power; The number of nodes in the output layer corresponds to the future time step to be predicted, and outputs the photovoltaic output power for the next 24 hours; The training set is input into the trained long short-term memory network, and the mean square error between the output photovoltaic output power for the next 24 hours and the actual output label is calculated. Adaptive estimation is used to adjust the weight parameters within the long short-term memory network according to the back propagation of the mean square error. The training set calculation process is iterated until the mean square error converges, and the generalization ability of the photovoltaic power generation model is tested using the validation set.

5. The method for dynamic optimization of a multi-scenario energy storage-photovoltaic-aluminum electrolysis system according to claim 1, characterized in that: The process of inputting the load characteristics and the factory production plan into the electrolytic aluminum load model and outputting the load baseline power and the adjustable flexibility range using the baseline and flexibility dichotomy method includes: Taking the load baseline level in the load characteristic as the load baseline power; The distribution range of the high-frequency disturbance component characteristics during normal production is statistically analyzed to obtain an initial adjustment range driven by historical data. The initial adjustment range is compared with a rigid safety boundary set according to the factory production plan and the electrolytic cell process requirements, and the intersection of the initial adjustment range and the rigid safety boundary is taken as the output adjustable flexible range.

6. A method for dynamic optimization of an energy storage-photovoltaic-aluminum electrolysis system adapted to multiple scenarios according to claims 2, 4 and 5, characterized in that: The process of obtaining a power forecast report by using the output of the photovoltaic power generation model and the output of the electrolytic aluminum load model includes: The system state prediction model calls the photovoltaic output power time series for the next 24 hours output by the photovoltaic power generation model and calls the load baseline power output by the electrolytic aluminum load model, and extrapolates the most recent load baseline power to the next 24 hours to form a load baseline power time series for the next 24 hours; The load baseline power time series and the photovoltaic output power time series are calculated in aligning manner, and the photovoltaic output power time series is subtracted from the load baseline power time series to obtain the net load forecast sequence of the energy storage-photovoltaic-electrolytic aluminum system in the next 24 hours. The photovoltaic output power time series, the load baseline power time series and the net load forecast sequence are integrated into a power forecast report.

7. A method for dynamic optimization of an energy storage-photovoltaic-aluminum electrolysis system adapted to multiple scenarios according to claims 2, 3 and 6, characterized in that: The process of outputting the day-ahead optimal economic dispatch plan by using the economic dispatch optimization model, based on the electricity price characteristics, the power forecast report, and the output of the energy storage system model, and adopting mixed integer linear programming includes: The economic dispatch optimization model uses mixed integer linear programming based on the grid exchange power, the real-time time-of-use electricity price provided by the time-of-use electricity price table, and the battery cycle aging cost, and optimizes and solves the problem with the minimum daily operating cost of the energy storage-photovoltaic-aluminum electrolysis system as the objective function; Establishing power balance constraints and energy storage system operation constraints using the charging power, the discharging power, and the binary variable that enables charging and discharging to occur at different times as decision variables; The power balance constraint is used to balance power generation and consumption of the energy storage system; The energy storage system operation constraints include a state of charge update constraint, a state of charge boundary constraint, and a charge and discharge power constraint. The state of charge update constraint is used to update the state of charge of the energy storage system according to the charge and discharge power at any time. The state of charge boundary constraint is used to preset the state of charge of the energy storage system between a minimum value and a maximum value. The charge and discharge power constraint is used to limit the charge and discharge power of the energy storage system to a preset maximum value range, and charging and discharging cannot be performed simultaneously. The objective function, the decision variables, and the operating constraints of the energy storage system are integrated, and the decision variable sequence that minimizes the objective function value is calculated through mixed integer linear programming. The decision variable sequence that minimizes the objective function value is used as the optimal charge and discharge power of the energy storage system at each time step in the next 24 hours, and the optimal charge and discharge power at each time step in the next 24 hours is formatted as a day-ahead optimal economic dispatch plan.

8. A dynamic optimization method for a multi-scenario energy storage-photovoltaic-aluminum electrolysis system according to claims 1 and 7, characterized in that: The process of outputting energy storage power control instructions by using the adaptive control model, based on the pre-processed photovoltaic system data, energy storage system data, electrolytic aluminum load power data and the day-ahead optimal economic dispatch plan, and using deep reinforcement learning technology includes: The state, action, and reward required for the adaptive control model to make the energy storage power control instruction decision are defined, wherein the state includes the real-time photovoltaic power of the photovoltaic system, the real-time electrolytic aluminum load power, and the power deviation from the day-ahead optimal economic dispatch plan at the current moment; the action is the energy storage power control instruction for the next time step, whose value range is constrained by the rated power of the energy storage converter; and the reward is a negative value related to the real-time operating cost; The energy storage system model, the photovoltaic power generation model, and the electrolytic aluminum load model together constitute a digital twin system, an intelligent agent is trained in the digital twin system, the intelligent agent outputs the energy storage power control instruction, and the digital twin system calculates the state and the reward at the next moment and feeds back to the intelligent agent; The agent is trained using a deep deterministic policy gradient algorithm. The agent performs simulated interactions in the digital twin system. In each simulated interaction, the agent selects an action based on the current state. The digital twin system feeds back the reward and the new state. The agent adjusts the weight parameters of its internal neural network based on the reward signal. The agent training process is iterated repeatedly until the agent's strategy converges and the maximum cumulative reward is continuously obtained. The intelligent agent strategy network that has been matured through offline training is deployed in the energy management system. At each time step of actual operation, the energy management system inputs the state data collected in real time into the intelligent agent strategy network. The intelligent agent strategy network outputs the optimal energy storage power control instruction under the current state through a forward propagation calculation.

9. A dynamic optimization method for a multi-scenario energy storage-photovoltaic-aluminum electrolysis system according to claims 1 and 8, characterized in that: The multiple scenarios include normal scenarios, extreme weather scenarios, and power grid failure scenarios. The process of combining the power grid data with the weather data and transmitting the energy storage power control instruction to the energy storage converter for execution after multi-scenario logic determination includes: When the grid data is within the normal operating range and the energy management system does not receive an external extreme weather warning signal, the current scenario is determined to be the normal scenario, and the energy storage power control instruction output by the adaptive control model is directly transmitted to the energy storage converter for execution; When the energy management system receives an external warning based on the meteorological data, it determines that the current scenario is the extreme meteorological scenario, temporarily suspends the energy storage power control instruction output by the adaptive control model, executes the emergency plan, generates a safety-oriented energy storage power control instruction, and transmits the safety-oriented energy storage power control instruction to the energy storage converter for execution; When the measurement, control and protection device of the grid connection point detects an abnormality in the grid data, it determines that the current scenario is the grid failure scenario, triggers the highest priority protection and control logic, issues an instruction to disconnect the circuit breaker connected to the public grid, and puts the plant microgrid into island operation mode. The energy storage power control instruction output by the adaptive control model is terminated, and the power supply control logic of the island operation is switched to maintain the voltage and frequency stability of the grid in the island, giving priority to ensuring continuous power supply to the core load of electrolytic aluminum. A balanced energy storage power control instruction is generated by the island operation controller according to the real-time power balancing demand, and the balanced energy storage power control instruction is transmitted to the energy storage converter.

10. A dynamic optimization method for a multi-scenario energy storage-photovoltaic-aluminum electrolysis system according to claims 1 and 9, characterized in that: The process of visualizing the energy storage charge state, the photovoltaic power, and the photovoltaic and load performance curve data through the integrated digital signage, and feeding back the execution effect of the energy storage power control instruction to the adaptive control model includes: The front-end application of the energy management system requests and aggregates the energy storage state of charge, photovoltaic power, and photovoltaic and load performance curve data from historical and real-time databases, and then structures and visualizes them on the human-computer interaction interface through dynamic charts, dashboards, and status indicators; After each energy storage power control instruction is executed, the energy management system records an experience data unit including the state before execution, the executed action, the actual reward and the new state after execution, and stores the experience data unit in an experience replay database. Data is regularly extracted from the experience replay database to perform offline retraining and optimization on the policy network of the adaptive control model, and the updated policy network is redeployed back to the energy management system.

Images