Power coordination management method and system based on virtual power plant concept in shelter hospital

Abstract

The invention provides an electric power coordination management method and system based on a virtual power plant concept in a shelter hospital. The method comprises the following steps: processing received real-time power demand information of each electric device in a shelter hospital and power generation prediction information of a renewable energy power generation unit, and generating optimized power demand and power generation prediction information; the working mode of each electric device and the charging and discharging strategy of the energy storage system are dynamically adjusted, the solving process of the model is optimized, and a globally optimal power dispatching scheme is generated; a price fluctuation rule of the electricity market is automatically learned, an electricity purchasing and selling strategy is calculated, a decision process is optimized, and an optimal electricity transaction strategy is generated; and when an abnormal condition is detected, analyzing and quickly positioning a fault reason, automatically starting a standby power supply, adjusting a power distribution plan, and reporting abnormal information to a management platform. According to the technical scheme provided by the invention, the reliability, the economical efficiency and the operation efficiency of the shelter hospital power system are improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of smart grid and energy management technology, and in particular to a method and system for coordinated power management in a square cabin hospital based on the concept of a virtual power plant. Background Art

[0002] As a temporary medical facility, Fangcang shelter hospitals are usually deployed quickly in emergency situations to provide necessary medical services. The power systems of such facilities need to be highly reliable and flexible to ensure the continuous operation of critical medical equipment. Since Fangcang shelter hospitals may be located in remote areas or emergency environments, their power supply often relies on renewable energy and energy storage systems. Therefore, how to efficiently manage and dispatch these distributed energy resources while ensuring a balance between power supply and demand and reducing operating costs has become an important technical requirement.

[0003] At present, the power management of Fangcang Hospital mainly relies on traditional power dispatching methods, including: formulating fixed power consumption plans based on historical data, but this method cannot cope with real-time changes in power demand and power generation. Relying on operators to manually monitor and adjust the working mode of each power-consuming equipment and the charging and discharging strategy of the energy storage system, but this method is inefficient and prone to errors. Automatically controlling the charging and discharging of the energy storage system based on preset rules, but this cannot fully consider the complex power supply and demand situation and market electricity price fluctuations.

[0004] However, most existing solutions adopt static or semi-static methods, which cannot respond to changes in electricity demand and the uncertainty of renewable energy generation in real time. This leads to an imbalance in electricity supply and demand, especially when weather conditions change or emergencies occur. Traditional methods fail to make full use of market electricity price information for optimization, resulting in uneconomical electricity purchase and sales strategies and increased operating costs. The existing power management system lacks a rapid response mechanism to abnormal situations. Once a failure occurs, it is difficult to quickly locate the cause and start the backup power supply, affecting the continuous power supply of critical medical equipment. The existing data recording and reporting mechanism may have security risks and cannot guarantee the security and immutability of data, affecting subsequent analysis and decision-making.

[0005] In summary, in order to improve the reliability and economy of the square cabin hospital power system, a more intelligent and dynamic power coordination management method is needed, which can process power demand and power generation forecast information in real time, optimize power dispatch, and further reduce costs through market electricity price information, while ensuring the security and non-tamperability of data. Summary of the invention

[0006] The embodiments of the present application provide a method and system for coordinated power management in a shelter hospital based on the concept of a virtual power plant, so as to solve the problems of low reliability, poor economy and low operating efficiency of the power system of the shelter hospital in the prior art.

[0007] In a first aspect, an embodiment of the present application provides a power coordination management method based on the concept of a virtual power plant in a square cabin hospital, including:

[0008] Analyze historical electricity consumption patterns and weather data, process the real-time power demand information of each electrical device in the square cabin hospital and the power generation forecast information of renewable energy power generation units, improve the accuracy of power generation forecast, evaluate the uncertainty of power demand, and generate optimized power demand and power generation forecast information;

[0009] Based on the optimized power demand and power generation forecast information, combined with the current state of the energy storage system in the square cabin hospital, a mixed integer linear programming model is used to dynamically adjust the working mode of each power-consuming equipment and the charging and discharging strategy of the energy storage system to ensure the balance of power supply and demand and maximize the use of renewable energy. A genetic algorithm is used to optimize the solution process of the mixed integer linear programming model to generate a globally optimal power dispatching plan;

[0010] According to the power dispatching scheme, market electricity price information is introduced, and the price fluctuation law of the power market is automatically learned through the reinforcement learning algorithm, and the power purchase and sale strategies are calculated to minimize the power cost. The decision-making process is further optimized through the simulated annealing algorithm to reduce the risk of power trading and generate the optimal power trading strategy;

[0011] Based on the optimal power trading strategy, the operating status of the cabin hospital's power system is monitored. When an abnormal situation is detected, the cause of the fault is analyzed and quickly located, the backup power supply is automatically started, and the power distribution plan is adjusted to ensure the continuous power supply of key medical equipment. The abnormal information is recorded to ensure the security and non-tamperability of the data, and the abnormal information is reported to the management platform so that timely response measures can be taken.

[0012] Optionally, based on the optimized power demand and power generation forecast information, combined with the current state of the energy storage system in the square cabin hospital, a mixed integer linear programming model is used to dynamically adjust the working mode of each power-consuming device and the charging and discharging strategy of the energy storage system to ensure the balance of power supply and demand and maximize the use of renewable energy, and a genetic algorithm is used to optimize the solution process of the mixed integer linear programming model to generate a globally optimal power dispatching plan, including:

[0013] Using the optimized power demand and power generation forecast information, preliminarily matching the demand of power-consuming equipment with the power generation capacity of renewable energy sources is performed to obtain a preliminary balance of power supply and demand;

[0014] Based on the preliminary power supply and demand balance, evaluate the current status of the energy storage system, including parameters of remaining power, maximum charging and discharging rates, and generate an energy storage system status report;

[0015] Based on the energy storage system status report and the preliminary power supply and demand balance, a mixed integer linear programming model is established with the goal of maximizing the utilization rate of renewable energy and the economy of the power system, wherein the variables include the working mode of the power equipment and the charging and discharging state of the energy storage system, and the constraints include power demand, power generation capacity and energy storage system capacity, to generate a mixed integer linear programming model;

[0016] Utilizing the mixed integer linear programming model, a genetic algorithm is used to optimize the model solution process to overcome the problem of local optimal solution and generate an optimized solution result;

[0017] Based on the optimized solution, a globally optimal power dispatching plan is generated to ensure the balance of power supply and demand, while minimizing power costs and improving the flexibility and reliability of the system.

[0018] Optionally, based on the energy storage system status report and the preliminary power supply and demand balance, a mixed integer linear programming model is established with the goal of maximizing the utilization rate of renewable energy and the economy of the power system, wherein the variables include the working mode of the power consumption equipment and the charging and discharging state of the energy storage system, and the constraints include power demand, power generation capacity and energy storage system capacity, and the mixed integer linear programming model is generated, including:

[0019] Using the energy storage system status report and the preliminary balance of power supply and demand, the model objective is determined to maximize the utilization rate of renewable energy and the economy of the power system, that is, to give priority to the use of renewable energy and reduce the cost of purchasing electricity from the external power grid while meeting all power demands;

[0020] Based on the model objectives, the working mode of the electric equipment and the charging and discharging state of the energy storage system are defined to generate model variables;

[0021] According to the model variables, the power demand, power generation capacity and energy storage system capacity are set to generate model constraints;

[0022] Using the model objectives, model variables and constraints, a mathematical expression is constructed to form a linear equation system;

[0023] Based on the linear equations, all elements are integrated to generate a complete mixed integer linear programming model as the basis of the optimization algorithm to find the best solution.

[0024] Optionally, based on the energy storage system status report and the preliminary power supply and demand balance, a mixed integer linear programming model is established with the goal of maximizing the utilization rate of renewable energy and the economy of the power system, wherein the variables include the working mode of the power consumption equipment and the charging and discharging state of the energy storage system, and the constraints include power demand, power generation capacity and energy storage system capacity, and the mixed integer linear programming model is generated, including:

[0025] Using the energy storage system status report and the preliminary power supply and demand balance, the model objective is determined to maximize the utilization rate of renewable energy and the economy of the power system, that is, on the premise of meeting all power demands, give priority to the use of renewable energy and reduce the cost of purchasing electricity from the external power grid. The objective function minZ is:

[0026]

[0027] Among them, C buy is the cost of purchasing a unit of electricity from the external power grid; P buy，t is the amount of electricity purchased from the external power grid at time t; P RE，t is the renewable energy power generation at time t; α is the weight coefficient of renewable energy utilization rate;

[0028] Based on the model objectives, the working mode of the electrical equipment and the charging and discharging state of the energy storage system are defined to generate model variables;

[0029] According to the model variables, the power demand, power generation capacity and energy storage system capacity are set to generate the constraints of the model. The following is a set of linear equations of the constraints:

[0030]

[0031] Among them, P load，t is the electricity demand at time t; E max is the maximum capacity of the energy storage system; P charge，max is the maximum charging rate of the energy storage system; P discharge，max is the maximum discharge rate of the energy storage system; P i，t is the power consumption of electrical equipment i at time t; η charge is the charging efficiency of the energy storage system; η discharge is the discharge efficiency of the energy storage system; x i，t is the working mode of the electrical equipment i at time t; P charge，t is the charging power of the energy storage system at time t; P discharge，t is the discharge power of the energy storage system at time t; E storage，t is the remaining power of the energy storage system at time t; It means that for all time points t, the relevant equations, inequalities or constraints must be valid at every time point t;

[0032] Based on the linear equations, all elements are integrated to generate a complete mixed integer linear programming model as the basis of the optimization algorithm to find the best solution.

[0033] Optionally, the mixed integer linear programming model is used to optimize the solution process of the model using a genetic algorithm to overcome the problem of local optimal solutions and generate an optimized solution result, including:

[0034] Using the mixed integer linear programming model, the population is initialized and a set of initial solutions are randomly generated, each solution represents a possible power dispatching scheme, covering the working mode of the power equipment and the charging and discharging state of the energy storage system;

[0035] Based on the initial solution, the fitness of the initial solution is evaluated, and the fitness value of the initial solution is calculated to reflect the performance of the initial solution in meeting power demand, maximizing renewable energy utilization, and minimizing power cost;

[0036] According to the fitness value, some excellent solutions are selected to enter the next generation population, and a roulette wheel selection or a tournament selection method is adopted to obtain the selected solution;

[0037] Utilizing the selected solutions, pairing and generating new solutions through crossover operations, simulating the gene recombination process in biological genetics, and obtaining a new solution population;

[0038] Based on the new solution population, a mutation operation is performed on some offspring solutions to randomly change the values ​​of certain variables, increase the diversity of the population, and generate a mutated solution population;

[0039] Using the mutated solution population, re-performing the fitness evaluation, selection, crossover and mutation operations on the new generation population, and continuously iterating until a stop condition is met, such as reaching a predetermined number of iterations or the fitness value no longer significantly increases;

[0040] Based on the solutions generated in all iterative processes, the solution with the highest fitness is selected as the global optimal solution to generate the optimized solution result.

[0041] Optionally, according to the power dispatching scheme, market electricity price information is introduced, the price fluctuation law of the power market is automatically learned through a reinforcement learning algorithm, the power purchase and sale strategies are calculated to minimize the power cost, and the decision-making process is further optimized through a simulated annealing algorithm to reduce the risk of power trading and generate the optimal power trading strategy, including:

[0042] Using the power dispatching scheme, introducing market electricity price information, collecting current and historical market electricity price data, including electricity price fluctuations in different time periods, and electricity price differences between peak and valley periods, to generate a market electricity price data set;

[0043] Based on the market electricity price data set, a reinforcement learning model is established to automatically learn the price fluctuation rules of the electricity market through continuous trial and error, gradually optimize the electricity purchase and sales strategy, and generate a preliminary electricity trading strategy;

[0044] The reinforcement learning model is trained using the market electricity price data set and the power dispatching plan. The reinforcement learning model will try different strategies for purchasing and selling electricity, provide feedback based on actual results, gradually adjust the strategy to minimize electricity costs, and generate an optimized electricity trading strategy.

[0045] Based on the optimized power trading strategy, the simulated annealing algorithm is used to optimize the decision-making process. By simulating the physical annealing process, the reinforcement learning model is helped to jump out of the local optimal solution and find the global optimal solution, thereby reducing the risk of power trading and generating a further optimized power trading strategy.

[0046] The optimization results of the comprehensive reinforcement learning model and the further optimization of the simulated annealing algorithm are used to generate the optimal electricity trading strategy to minimize the electricity cost while maintaining high flexibility and adaptability and reducing trading risks.

[0047] Optionally, according to the power dispatching scheme, market electricity price information is introduced, the price fluctuation law of the power market is automatically learned through a reinforcement learning algorithm, the power purchase and sale strategies are calculated to minimize the power cost, and the decision-making process is further optimized through a simulated annealing algorithm to reduce the risk of power trading and generate the optimal power trading strategy, including:

[0048] Using the power dispatching scheme, introducing market electricity price information, collecting current and historical market electricity price data, including electricity price fluctuations in different time periods, and electricity price differences between peak and valley periods, to generate a market electricity price data set;

[0049] Based on the market electricity price data set, a reinforcement learning model is established to automatically learn the price fluctuation rules of the electricity market through continuous trial and error, gradually optimize the electricity purchase and sales strategies, and generate a preliminary electricity trading strategy, as follows:

[0050] Initializing a reinforcement learning model, wherein the state space of the reinforcement learning model includes the current market electricity price, electricity demand, and energy storage system status, and the action space includes decisions on electricity purchase and sale;

[0051] Define a reward function to evaluate the performance of the reinforcement learning model in each decision step. The reward function R(t) is:

[0052]

[0053] Among them, C buy (t) is the cost of purchasing a unit of electricity from the external power grid at time t; P buy (t) is the amount of electricity purchased from the external power grid at time t; P RE (t) is the amount of electricity generated by renewable energy at time t; P discharge (t) is the discharge power of the energy storage system at time t; P load (t) is the electricity demand at time t; β is the weight coefficient of the electricity supply and demand balance; γ is the exponential parameter of the supply and demand balance, which is used to adjust the importance of the supply and demand balance; ∈ is a small constant to prevent division by zero errors;

[0054] The reinforcement learning model is trained using the market electricity price data set and the power dispatching plan. The reinforcement learning model will try different strategies for purchasing and selling electricity, provide feedback based on actual results, gradually adjust the strategy to minimize electricity costs, and generate an optimized electricity trading strategy.

[0055] Based on the optimized power trading strategy, the simulated annealing algorithm is used to optimize the decision-making process. By simulating the physical annealing process, the reinforcement learning model is helped to jump out of the local optimal solution, find the global optimal solution, reduce the risk of power trading, and generate a further optimized power trading strategy, as follows:

[0056] Initialize the simulated annealing algorithm and set the initial temperature T 0 , cooling rate α, termination temperature T final ;

[0057] Define the energy function to evaluate the quality of the current electricity trading strategy. The energy function E(t) is:

[0058]

[0059] Among them, δ is the weight coefficient of power supply and demand balance; η is the index parameter of supply and demand balance, which is used to adjust the importance of supply and demand balance;

[0060] Performing a neighborhood search on the optimized power trading strategy to generate a set of neighboring candidate solutions;

[0061] According to the energy function, the energy values ​​of the current solution and the candidate solution are calculated. If the energy value of the candidate solution is lower, the candidate solution is accepted as the new current solution; if the energy value of the candidate solution is higher, the candidate solution is accepted with a certain probability. The probability decreases as the temperature decreases. The specific probability formula is:

[0062]

[0063] Among them, E new is the energy value of the candidate solution; E current is the energy value of the current solution; T is the current temperature; P accept represents the probability of accepting a candidate solution;

[0064] Using the updated current solution, the temperature is gradually reduced according to the preset cooling rate to help the reinforcement learning model jump out of the local optimal solution, explore a larger solution space, and generate the simulated annealing algorithm state after cooling. The temperature update formula is:

[0065] T new =α·T current

[0066] Among them, T new is the new temperature; T current is the current temperature; α is the cooling rate; or the temperature reduction coefficient

[0067] When the temperature drops to the termination temperature or reaches a predetermined number of iterations, the simulated annealing algorithm is stopped, and a further optimized power trading strategy is generated based on the simulated annealing algorithm state after the temperature drops;

[0068] By combining the optimization results of the reinforcement learning model and the further optimization of the simulated annealing algorithm, an optimal electricity trading strategy is generated to minimize electricity costs while maintaining high flexibility and adaptability and reducing trading risks.

[0069] Optionally, based on the market electricity price data set, a reinforcement learning model is established to automatically learn the price fluctuation rules of the electricity market through continuous trial and error, gradually optimize the electricity purchase and sale strategies, and generate a preliminary electricity trading strategy, including:

[0070] Using the market electricity price data set, a reinforcement learning model is initialized, wherein the state space of the reinforcement learning model includes the current market electricity price, electricity demand, and energy storage system state, and the action space includes decisions on electricity purchase and sale, thereby generating an initialized reinforcement learning model;

[0071] Based on the market electricity price data set and the power dispatch plan, a reward function is defined to evaluate the performance of the reinforcement learning model in each decision step. Common reward indicators include the level of electricity cost and the degree of balance between electricity supply and demand, and a reward function is generated;

[0072] The initialized reinforcement learning model is trained using the market electricity price data set, the power dispatching plan and the reward function. The reinforcement learning model will try different strategies for purchasing and selling electricity, provide feedback based on actual results, and gradually adjust the strategy to minimize electricity costs, thereby generating a trained reinforcement learning model.

[0073] Based on the trained reinforcement learning model, a preliminary electricity trading strategy is generated, which includes electricity purchase and sale decisions in different time periods, aiming to minimize electricity costs and ensure a balance between electricity supply and demand.

[0074] Optionally, based on the optimized power trading strategy, a simulated annealing algorithm is used to optimize the decision-making process, and by simulating a physical annealing process, the reinforcement learning model is helped to jump out of the local optimal solution and find the global optimal solution, thereby reducing the risk of power trading and generating a further optimized power trading strategy, including:

[0075] Using the optimized power trading strategy, the simulated annealing algorithm is initialized, the parameters of the initial temperature, the cooling rate, and the termination temperature are set, and the initialized simulated annealing algorithm is generated;

[0076] Based on the optimized power trading strategy, an energy function is defined to evaluate the quality of the current power trading strategy. The energy function includes indicators of power cost and the degree of balance between power supply and demand. The lower the energy value, the better the strategy. The energy function is generated.

[0077] Using the initialized simulated annealing algorithm and the energy function, a neighborhood search is performed on the optimized power trading strategy to generate a set of neighboring candidate solutions;

[0078] According to the energy function, the energy values ​​of the current solution and the candidate solution are calculated. If the energy value of the candidate solution is lower, the candidate solution is accepted as the new current solution; if the energy value of the candidate solution is higher, the candidate solution is accepted with a certain probability, and the probability decreases as the temperature decreases, to generate an updated current solution;

[0079] Using the updated current solution, gradually reducing the temperature according to a preset cooling rate, helping the reinforcement learning model to jump out of the local optimal solution, explore a larger solution space, and generate a simulated annealing algorithm state after cooling;

[0080] When the temperature drops to the termination temperature or reaches a predetermined number of iterations, the simulated annealing algorithm is stopped, and based on the state of the simulated annealing algorithm after the temperature drop, a further optimized electricity trading strategy is generated to minimize electricity costs while maintaining high flexibility and adaptability and reducing trading risks.

[0081] In a second aspect, an embodiment of the present application provides a power coordination management system based on the concept of a virtual power plant in a square cabin hospital, including:

[0082] The analysis and evaluation module is used to analyze historical power consumption patterns and weather data, process the real-time power demand information of each power-consuming device in the square cabin hospital and the power generation forecast information of the renewable energy power generation unit, improve the accuracy of power generation forecast, evaluate the uncertainty of power demand, and generate optimized power demand and power generation forecast information;

[0083] An adjustment and optimization module is used to dynamically adjust the working mode of each power-consuming device and the charging and discharging strategy of the energy storage system based on the optimized power demand and power generation forecast information, combined with the current state of the energy storage system in the square cabin hospital, using a mixed integer linear programming model to ensure the balance of power supply and demand and maximize the use of renewable energy, and use a genetic algorithm to optimize the solution process of the mixed integer linear programming model to generate a globally optimal power dispatching plan;

[0084] A calculation generation module is used to introduce market electricity price information according to the power dispatch plan, automatically learn the price fluctuation law of the power market through the reinforcement learning algorithm, calculate the power purchase and sale strategy, minimize the power cost, and further optimize the decision-making process through the simulated annealing algorithm, reduce the risk of power trading, and generate the optimal power trading strategy;

[0085] The monitoring and reporting module is used to monitor the operating status of the power system of the cabin hospital based on the optimal power trading strategy. When an abnormal situation is detected, the cause of the fault is quickly located by analysis, the backup power supply is automatically started, and the power distribution plan is adjusted to ensure the continuous power supply of key medical equipment. The abnormal information is recorded to ensure the security and non-tamperability of the data, and the abnormal information is reported to the management platform so that timely response measures can be taken.

[0086] In the embodiment of the present application, historical electricity consumption patterns and weather data are analyzed, and the received real-time power demand information of each electrical equipment in the square cabin hospital and the power generation forecast information of the renewable energy power generation unit are processed to improve the accuracy of the power generation forecast, and the uncertainty of the power demand is evaluated to generate optimized power demand and power generation forecast information; based on the optimized power demand and power generation forecast information, combined with the current state of the energy storage system in the square cabin hospital, a mixed integer linear programming model is used to dynamically adjust the working mode of each electrical equipment and the charging and discharging strategy of the energy storage system to ensure the balance of power supply and demand and maximize the use of renewable energy, and a genetic algorithm is used to optimize the solution process of the mixed integer linear programming model to generate a global optimal power regulation According to the power dispatching plan, the market electricity price information is introduced, the price fluctuation law of the electricity market is automatically learned through the reinforcement learning algorithm, the power purchase and sale strategies are calculated to minimize the power cost, and the decision-making process is further optimized through the simulated annealing algorithm to reduce the risk of power trading and generate the optimal power trading strategy; based on the optimal power trading strategy, the operating status of the power system of the cabin hospital is monitored. When an abnormal situation is detected, the cause of the fault is analyzed and quickly located, the backup power supply is automatically started, and the power distribution plan is adjusted to ensure the continuous power supply of key medical equipment, and the abnormal information is recorded to ensure the security and non-tamperability of the data, and the abnormal information is reported to the management platform so that timely countermeasures can be taken.

[0087] The method proposed in the embodiment of the present application comprehensively analyzes historical electricity consumption patterns and weather data, processes electricity demand and power generation forecast information in real time, and dynamically adjusts the working mode of each power-consuming equipment and the charging and discharging strategy of the energy storage system in combination with the status of the energy storage system, thereby ensuring the balance of electricity supply and demand and maximizing the use of renewable energy. At the same time, it optimizes the electricity trading strategy through market electricity price information, thereby improving the reliability, economy and operation efficiency of the electric power system of the square cabin hospital.

[0088] Furthermore, the method proposed in the embodiment of the present application preliminarily matches electricity demand with renewable energy generation capacity, evaluates the state of the energy storage system, establishes a mixed integer linear programming model, and uses a genetic algorithm to optimize the solution process to generate a globally optimal power dispatching plan, thereby ensuring the balance of electricity supply and demand, minimizing electricity costs, and improving the flexibility and reliability of the system.

[0089] Furthermore, the method proposed in the embodiment of the present application introduces market electricity price information, uses a reinforcement learning algorithm to automatically learn the price fluctuation patterns of the electricity market, gradually optimizes electricity purchasing and selling strategies, and further optimizes the decision-making process through a simulated annealing algorithm, thereby reducing the risk of electricity trading, generating an optimal electricity trading strategy, and minimizing electricity costs while maintaining high flexibility and adaptability.

[0090] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0092] Figure 1 A flowchart of a method for coordinated power management based on the concept of a virtual power plant in a square cabin hospital provided in an embodiment of the present application;

[0093] Figure 2 A structural diagram of an electric power coordination and management system based on the concept of a virtual power plant in a square cabin hospital provided in an embodiment of the present application. DETAILED DESCRIPTION

[0094] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0095] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0096] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0097] Figure 1 A flowchart of a power coordination management method based on the concept of a virtual power plant in a square cabin hospital is provided for an embodiment of the present application, such as Figure 1 As shown, the method includes:

[0098] 101. Analyze historical electricity consumption patterns and weather data, process the real-time power demand information of each electrical device in the square cabin hospital and the power generation forecast information of the renewable energy power generation unit, improve the accuracy of the power generation forecast, evaluate the uncertainty of power demand, and generate optimized power demand and power generation forecast information;

[0099] In this step, the historical power consumption pattern refers to the actual power consumption of each electrical device in the Fangcang Hospital over the past period of time. Weather data includes meteorological parameters such as temperature, humidity, and wind speed, which affect the power generation of renewable energy. Real-time power demand information refers to the power demand of each electrical device in the Fangcang Hospital at the current moment. Renewable energy generation forecast information predicts the power generation of renewable energy generation units in the future based on weather data and other factors.

[0100] By analyzing historical electricity consumption patterns and weather data, combined with real-time electricity demand information and renewable energy generation forecast information, data analysis and machine learning techniques are used to improve the accuracy of power generation forecasts. At the same time, the uncertainty of electricity demand is assessed to generate more accurate and reliable electricity demand and generation forecast information.

[0101] In this application example, in the square cabin hospital, the system can collect electricity consumption records and weather data from the past year, and use the time series analysis model to predict the power demand in the next 24 hours. At the same time, combined with the current weather forecast and historical power generation data, the power generation of solar panels and wind turbines is predicted. Through this forecast information, the system can more accurately estimate the future power supply and demand, thereby providing a reliable basis for subsequent power dispatch.

[0102] 102. Based on the optimized power demand and power generation forecast information, combined with the current state of the energy storage system in the square cabin hospital, a mixed integer linear programming model is used to dynamically adjust the working mode of each power-consuming device and the charging and discharging strategy of the energy storage system to ensure the balance of power supply and demand and maximize the use of renewable energy, and a genetic algorithm is used to optimize the solution process of the mixed integer linear programming model to generate a globally optimal power dispatching plan;

[0103] In this step, the mixed integer linear programming model is a mathematical optimization model used to solve linear programming problems with discrete decision variables. The energy storage system includes batteries or other forms of energy storage devices to store excess electrical energy and release it when needed. The genetic algorithm is a heuristic search algorithm that simulates natural selection and genetic mechanisms to find approximate optimal solutions to complex optimization problems.

[0104] Based on the optimized power demand and power generation forecast information, combined with the current state of the energy storage system, a mixed integer linear programming model is established. This model ensures the balance of power supply and demand and maximizes the use of renewable energy by adjusting the working mode of each power device and the charging and discharging strategy of the energy storage system. In order to overcome the problem of local optimal solution, the genetic algorithm is used to optimize the solution process of the model and generate the global optimal power dispatch plan.

[0105] In this application example, it is assumed that there are multiple medical devices and a large energy storage system in the square cabin hospital. The system establishes a mixed integer linear programming model based on the predicted power demand and renewable energy generation, combined with the current state of the energy storage system. The model takes into account the operating constraints of each device and the charging and discharging limitations of the energy storage system. Through iterative optimization of the genetic algorithm, a global optimal power dispatching plan is finally generated to ensure the balance of power supply and demand while maximizing the use of renewable energy.

[0106] Optionally, in step 102, based on the optimized power demand and power generation forecast information, combined with the current state of the energy storage system in the square cabin hospital, a mixed integer linear programming model is used to dynamically adjust the working mode of each power device and the charging and discharging strategy of the energy storage system to ensure the balance of power supply and demand and maximize the use of renewable energy, and a genetic algorithm is used to optimize the solution process of the mixed integer linear programming model to generate a globally optimal power dispatching plan, including: using the optimized power demand and power generation forecast information to preliminarily match the demand of power devices with the power generation capacity of renewable energy to obtain a preliminary power supply and demand balance; based on the preliminary power supply and demand balance, the current state of the energy storage system is evaluated, including the remaining power, maximum charge and The invention relates to a method for generating a mixed integer linear programming model for maximizing the utilization rate of renewable energy and the economy of the power system based on the parameters of the discharge rate; a mixed integer linear programming model is established based on the energy storage system status report and the preliminary balance of power supply and demand, wherein the variables include the working mode of the power-consuming equipment and the charge and discharge state of the energy storage system, and the constraints include power demand, power generation capacity and energy storage system capacity, to generate a mixed integer linear programming model; the mixed integer linear programming model is used to optimize the solution process of the model using a genetic algorithm to overcome the problem of local optimal solutions and generate an optimized solution result; based on the optimized solution result, a global optimal power dispatching plan is generated to ensure the balance of power supply and demand, while minimizing power costs and improving the flexibility and reliability of the system.

[0107] Optionally, in step 102, based on the energy storage system status report and the preliminary power supply and demand balance, a mixed integer linear programming model is established with the goal of maximizing the utilization rate of renewable energy and the economy of the power system, wherein the variables include the working mode of the power equipment and the charging and discharging state of the energy storage system, and the constraints include power demand, power generation capacity and energy storage system capacity, and the mixed integer linear programming model is generated, including: using the energy storage system status report and the preliminary power supply and demand balance, determining that the model goal is to maximize the utilization rate of renewable energy and the economy of the power system, that is, on the premise of meeting all power demands, give priority to the use of renewable energy and reduce the cost of purchasing electricity from the external power grid; based on the model goal, the working mode of the power equipment and the charging and discharging state of the energy storage system are defined to generate model variables; based on the model variables, the power demand, power generation capacity and energy storage system capacity are set to generate model constraints; using the model goal, model variables and constraints, a mathematical expression is constructed to form a linear equation group; based on the linear equation group, all elements are integrated to generate a complete mixed integer linear programming model as the basis of the optimization algorithm to find the best solution.

[0108] Optionally, the method of using the mixed integer linear programming model in step 102 to solve the model using a genetic algorithm to overcome the problem of local optimal solutions and generate optimized solution results includes: using the mixed integer linear programming model to initialize the population and randomly generate a set of initial solutions, each solution representing a possible power dispatching scheme, covering the working mode of the power equipment and the charging and discharging state of the energy storage system; based on the initial solution, performing fitness evaluation on the initial solution, calculating the fitness value of the initial solution, reflecting the performance of the initial solution in meeting power demand, maximizing the utilization rate of renewable energy and minimizing power cost; according to the fitness value, selecting some excellent solutions to enter the next generation population, and adopting A selected solution is obtained by a roulette wheel selection or a tournament selection method; the selected solution is used to perform pairing and generate a new solution through a crossover operation, simulating the gene recombination process in biological genetics to obtain a new solution population; based on the new solution population, a mutation operation is performed on part of the offspring solutions to randomly change the values ​​of certain variables, increase the diversity of the population, and generate a mutated solution population; using the mutated solution population, the new generation population is subjected to the fitness evaluation, selection, crossover and mutation operations again, and iterates continuously until a stopping condition is met, such as reaching a predetermined number of iterations or the fitness value is no longer significantly improved; based on the solutions generated in all iterative processes, the solution with the highest fitness is selected as the global optimal solution to generate an optimized solution result.

[0109] In this step, mixed integer linear programming is a mathematical optimization technique used to solve decision problems with linear objective functions and linear constraints, in which some variables must take integer values. In the square cabin hospital power dispatch scheme, the mixed integer linear programming model maximizes the utilization of renewable energy and reduces electricity costs by defining the working mode of power equipment and the charging and discharging status of the energy storage system as variables, and considering constraints such as power demand, power generation capacity and energy storage system capacity. Genetic algorithm is a search heuristic algorithm that simulates natural selection and genetic mechanisms. It searches for the global optimal solution through selection, crossover and mutation operations in an iterative process, overcoming the problem of local optimal solutions that may be encountered in the mixed integer linear programming solution process.

[0110] First, based on the optimized power demand and power generation forecast information, as well as the current state of the energy storage system, the demand for power equipment is preliminarily matched with the power generation capacity of renewable energy. Then, based on the preliminary matching results, the state of the energy storage system is evaluated, and a mixed integer linear programming model is established with the goal of maximizing the utilization rate and economy of renewable energy, which includes the variable definitions and related constraints of the working mode of power equipment and the charging and discharging strategy of the energy storage system. Next, the genetic algorithm is used to optimize and solve the mixed integer linear programming model: a set of random solutions is initialized as the initial population, and then the population is continuously iterated and updated through a series of steps such as fitness evaluation, selection, crossover, and mutation until the predetermined stop condition is met. Finally, the best solution is selected from all iterative solutions as the global optimal power dispatching solution.

[0111] In an embodiment of the present application, it is assumed that a square cabin hospital has solar panels as the main source of renewable energy and is equipped with a battery energy storage system. During the day, the hospital is expected to have different peak and valley power demand changes; at the same time, the weather forecast shows that it will experience a cloudy to sunny process in the next 24 hours, which means that photovoltaic power generation will fluctuate. In this context, the above scheme is applied to first estimate the hourly power demand and photovoltaic output based on historical data and weather forecast information. Then, combined with the current battery power level and its maximum charge and discharge rate, a mixed integer linear programming model is constructed to minimize the power purchase of the power grid while maximizing the efficiency of solar energy use. The model is optimized and solved using a genetic algorithm, and the optimal scheduling strategy is obtained after multiple iterations-for example, charging the battery first during periods of sufficient light and using solar power as much as possible, and reasonably arranging battery discharge at night or on cloudy days to supplement the insufficient power supply, so as to achieve a balance of power supply and demand throughout the cycle at the lowest cost.

[0112] This application considers that in the coordinated management of electricity in Fangcang Hospital, in order to ensure the balance of electricity supply and demand and maximize the use of renewable energy while reducing the cost of purchasing electricity from the external power grid, a mixed integer linear programming model needs to be established. The goal of this model is to give priority to the use of renewable energy while meeting all electricity needs, and to achieve the economy of the power system by optimizing the charging and discharging strategy of the energy storage system.

[0113] Optionally, in step 102, based on the energy storage system status report and the preliminary power supply and demand balance, a mixed integer linear programming model is established with the goal of maximizing the utilization rate of renewable energy and the economy of the power system, wherein the variables include the working mode of the power consumption equipment and the charging and discharging state of the energy storage system, and the constraints include power demand, power generation capacity and energy storage system capacity, and the mixed integer linear programming model is generated, including:

[0114] Using the energy storage system status report and the preliminary power supply and demand balance, the model objective is determined to maximize the utilization rate of renewable energy and the economy of the power system, that is, on the premise of meeting all power demands, give priority to the use of renewable energy and reduce the cost of purchasing electricity from the external power grid. The objective function minZ is:

[0115]

[0116] Among them, C buy is the cost of purchasing a unit of electricity from the external power grid; P buy，t is the amount of electricity purchased from the external power grid at time t; P RE，t is the renewable energy power generation at time t; α is the weight coefficient of renewable energy utilization rate;

[0117] Based on the model objectives, the working mode of the electrical equipment and the charging and discharging state of the energy storage system are defined to generate model variables;

[0118] According to the model variables, the power demand, power generation capacity and energy storage system capacity are set to generate the constraints of the model. The following is a set of linear equations of the constraints:

[0119]

[0120] Among them, P load，t is the electricity demand at time t; E max is the maximum capacity of the energy storage system; P charge，max is the maximum charging rate of the energy storage system; P discharge，max is the maximum discharge rate of the energy storage system; P i，t is the power consumption of electrical equipment i at time t; η charge is the charging efficiency of the energy storage system; η discharge is the discharge efficiency of the energy storage system; xi，t is the working mode of the electrical equipment i at time t; P charge，t is the charging power of the energy storage system at time t; P discharge，t is the discharge power of the energy storage system at time t; E storage，t is the remaining power of the energy storage system at time t; It means that for all time points t, the relevant equations, inequalities or constraints must be established at every time point t; t represents the time point, which can be a discrete time step, such as every hour, every minute, etc.;

[0121] Based on the linear equations, all elements are integrated to generate a complete mixed integer linear programming model as the basis of the optimization algorithm to find the best solution.

[0122] The objective function is designed to minimize the cost of purchasing electricity from the external power grid and maximize the utilization of renewable energy. The constraints are to ensure the balance of power supply and demand, the capacity limit of the energy storage system, the working mode of the power equipment and the charging and discharging status of the energy storage system.

[0123] The following is a brief introduction to the design reasons of each sub-item of the formula:

[0124] In the objective function formula, C buy ·P buy，t This part represents the cost of purchasing electricity from the external grid. By minimizing this cost, you can reduce your reliance on the external grid, thereby reducing operating costs. t P RE，t This part represents the weighted value of renewable energy generation. By maximizing this part, more renewable energy can be encouraged to be used, thus improving the sustainability and environmental friendliness of the system.

[0125] The following is a brief introduction to how to obtain the parameters of the formula:

[0126] Among them, C buy Real-time electricity price data can be obtained from the electricity market or suppliers. buy，t is the amount of electricity purchased from the external grid at each time point t, and is a decision variable of the model. α is a weight coefficient that can be set according to actual needs. For example, if more attention is paid to the utilization rate of renewable energy, a larger α value is set. RE，t is the amount of renewable energy power generation at each time point t, which can be predicted through weather forecasts and historical data.

[0127] The following is a brief introduction to the design reasons of each sub-item of the formula:

[0128] In the mixed integer linear programming model, this sub-item ∑ i x i，t ·Pi，t The reason for the design is to ensure that at each time point t, the sum of the actual power consumption of all electrical devices is equal to the total power demand P load，t , thereby ensuring a balance between electricity supply and demand.

[0129] The following is a brief introduction to how to obtain the parameters of the formula:

[0130] Among them, x i，t is the working state of the electrical equipment i at time point t (0 or 1), P i，t P is the power consumption of the power consumption device i at time t, which can be obtained from the device specifications and operation plan. load，t is the total electricity demand, which can be obtained through forecasting.

[0131] The following is a brief introduction to the design reasons of each sub-item of the formula:

[0132] In the energy update formula of the energy storage system, E storage，t Represents the remaining energy of the energy storage system at the current time point t. This is the known initial condition or the calculation result at the previous time point; η charge ·P charge，t Represents the additional energy added by the energy storage system at time t. Here the charging efficiency η is taken into account charge , because there will be energy loss in the actual charging process; It represents the energy reduction caused by the energy storage system discharging at time point t. The discharge efficiency η is considered here. discharge , because there will be energy loss in the actual discharge process.

[0133] The following is a brief introduction to how to obtain the parameters of the formula:

[0134] Among them, E storage，t It can be obtained from the status report of the energy storage system, or updated as a state variable during the operation of the model; η charge is the charging efficiency of the energy storage system, usually given in the equipment specification; P charge，t is the charging power of the energy storage system at time point t, which is the decision variable of the model; η discharge is the discharge efficiency of the energy storage system, usually given in the equipment specification; P discharge，t is the discharge power of the energy storage system at time point t and is the decision variable of the model.

[0135] The following is a brief introduction to the other parameter acquisition methods:

[0136] Among them, E max is the rated capacity of the energy storage system, obtained from the equipment specifications;

[0137] Assume that the Fangcang Hospital has the following parameters:

[0138] C buy =0.15 yuan / kWh; α = 10; time step per hour;

[0139] Table 1 below is the electricity demand and renewable energy generation forecast:

[0140] Table 1

[0141]

[0142] The following are the energy storage system parameters:

[0143] E max =200kWh; P charge，max =50kW; P discharge，max =50kW; η charge =0.9; η discharge =0.9;

[0144] The following is the initial energy storage system state: E storage，0 =100kWh

[0145] Table 2 below shows the working modes and power consumption of electrical equipment:

[0146] Table 2

[0147]

[0148]

[0149] Calculate the electricity supply and demand balance:

[0150] For each time point t, calculate P buy，t and P charge，t With P discharge，t ; Time point 0:

[0151] P RE，0 +P discharge，0 -P charge，0 +P buy，0 =P load，0

[0152] 50+P discharge，0 -P charge，0 +P buy，0 =100

[0153] Assume P discharge，0 =20kW, P charge，0 =10kW, then

[0154] 50+20-10+P buy，0 =100

[0155] P buy，0 =40kW

[0156] Time point 1:

[0157] P RE，1 +P discharge，1 -P charge，1 +P buy，1 =P load，1

[0158] 60+P discharge，1 -P charge，1 +P buy，1 =120

[0159] Assume P discharge，1 =30kW, P charge，1 =10kW, then

[0160] 60+30-10+P buy，1 =120

[0161] P buy，1 =40kW

[0162] Time point 2:

[0163] P RE，2 +P discharge，2 -P charge，2 +P buy，2 =P load，2

[0164] 40+P discharge，2 -P charge，2 +P buy，2 =110

[0165] Assume P discharge，2 =20kW, P charge，2 =0kW, then

[0166] 40+20-0+P buy，2 =110

[0167] P buy，2 =50kW

[0168] Calculate the energy storage system energy update:

[0169] Time point 0:

[0170]

[0171] Time point 1:

[0172]

[0173] Time point 2:

[0174]

[0175] Objective function calculation

[0176]

[0177] minZ=0.15·(40+40+50)-10·(50+60+40)

[0178] minZ=0.15·130-10·150

[0179] minZ=19.5-1500

[0180] minZ=-1480.5

[0181] The objective function value is -1480.5, which means that after considering the external power grid power purchase cost and renewable energy utilization rate, the overall cost is negative. This is mainly because the weight coefficient α of renewable energy utilization rate is large, which causes the contribution of renewable energy to be significantly amplified. The balance of power supply and demand balances the power supply and demand at each time point, ensuring the stable operation of the power system. The state of the energy storage system keeps the remaining energy of the energy storage system within a reasonable range at each time point, neither exceeding the maximum capacity nor falling below zero. By optimizing the charging and discharging strategy of the energy storage system, the external power grid power purchase cost is reduced, while the utilization rate of renewable energy is maximized.

[0182] This example demonstrates how to optimize the power dispatch of a shelter hospital through a mixed integer linear programming model to ensure a balance between power supply and demand and improve the economy and reliability of the system.

[0183] 103. According to the power dispatching plan, market electricity price information is introduced, and the price fluctuation law of the power market is automatically learned through the reinforcement learning algorithm, and the power purchase and sale strategies are calculated to minimize the power cost. The decision-making process is further optimized through the simulated annealing algorithm to reduce the risk of power trading and generate the optimal power trading strategy;

[0184] In this step, the market electricity price information refers to the electricity price data in different time periods in the electricity market. The reinforcement learning algorithm is a machine learning method that continuously learns and optimizes strategies through interaction with the environment. The simulated annealing algorithm is a global optimization algorithm that simulates the physical annealing process to help jump out of the local optimal solution.

[0185] Based on the power dispatching plan, the market electricity price information is introduced, and the price fluctuation law of the power market is automatically learned through the reinforcement learning algorithm, and the power purchase and sales strategy is gradually optimized to minimize the power cost. In order to further optimize the decision-making process, the simulated annealing algorithm is used to help jump out of the local optimal solution, reduce the risk of power trading, and generate the optimal power trading strategy.

[0186] In this application example, the power management system of the Fangcang Hospital collects current and historical market electricity price data and establishes a reinforcement learning model. The model gradually optimizes the power purchase and sale strategy through continuous trial and error learning. For example, more power is purchased when the power price is low, and excess power is sold when the power price is high. The decision-making process is further optimized through the simulated annealing algorithm to ensure that the power trading strategy is not only economical but also low-risk.

[0187] Optionally, the method in step 103 introduces market electricity price information according to the power dispatching scheme, automatically learns the price fluctuation law of the power market through the reinforcement learning algorithm, calculates the power purchase and sale strategy, minimizes the power cost, and further optimizes the decision-making process through the simulated annealing algorithm to reduce the risk of power trading and generate the optimal power trading strategy, including: using the power dispatching scheme to introduce market electricity price information, collect current and historical market electricity price data, including electricity price fluctuations in different time periods, and electricity price differences between peak and trough periods, and generate a market electricity price data set; based on the market electricity price data set, establishes a reinforcement learning model, automatically learns the price fluctuation law of the power market through continuous trial and error, gradually optimizes the electricity purchase and sale strategy, and generates a preliminary power trading strategy; using The reinforcement learning model is trained using the market electricity price data set and the power dispatching plan. The reinforcement learning model will try different strategies for purchasing and selling electricity, provide feedback based on actual results, and gradually adjust strategies to minimize electricity costs, thereby generating an optimized electricity trading strategy. Based on the optimized electricity trading strategy, a simulated annealing algorithm is used to optimize the decision-making process. By simulating the physical annealing process, the reinforcement learning model is helped to jump out of the local optimal solution and find the global optimal solution, thereby reducing the risk of electricity trading and generating a further optimized electricity trading strategy. The optimization results of the comprehensive reinforcement learning model and the further optimization of the simulated annealing algorithm are combined to generate an optimal electricity trading strategy, thereby minimizing electricity costs while maintaining high flexibility and adaptability and reducing trading risks.

[0188] Optionally, the method in step 103 is to establish a reinforcement learning model based on the market electricity price data set, automatically learn the price fluctuation law of the electricity market through continuous trial and error, gradually optimize the electricity purchase and sale strategy, and generate a preliminary electricity trading strategy, including: using the market electricity price data set to initialize a reinforcement learning model, the state space of the reinforcement learning model includes the current market electricity price, electricity demand, and energy storage system status, and the action space includes electricity purchase and sale decisions, to generate an initialized reinforcement learning model; based on the market electricity price data set and the power dispatch plan, define a reward function to evaluate the performance of the reinforcement learning model in each decision step , common reward indicators include the level of electricity cost and the degree of balance between electricity supply and demand, and a reward function is generated; the initialized reinforcement learning model is trained using the market electricity price data set, the power dispatch plan and the reward function, and the reinforcement learning model will try different electricity purchase and sale strategies, and provide feedback based on the actual results, and gradually adjust the strategy to minimize the electricity cost, and generate a trained reinforcement learning model; based on the trained reinforcement learning model, a preliminary electricity trading strategy is generated, and the preliminary electricity trading strategy includes electricity purchase and sale decisions in different time periods, aiming to minimize the electricity cost and ensure the balance between electricity supply and demand.

[0189] Optionally, in step 103, based on the optimized power trading strategy, a simulated annealing algorithm is used to optimize the decision-making process, and by simulating the physical annealing process, the reinforcement learning model is helped to jump out of the local optimal solution and find the global optimal solution, thereby reducing the risk of power trading and generating a further optimized power trading strategy, including: using the optimized power trading strategy, initializing the simulated annealing algorithm, setting the parameters of the initial temperature, cooling rate, and termination temperature, and generating an initialized simulated annealing algorithm; based on the optimized power trading strategy, defining an energy function for evaluating the quality of the current power trading strategy, the energy function includes indicators of power cost and the degree of balance between power supply and demand, the lower the energy value, the better the strategy, and generating an energy function; using the initialized simulated annealing algorithm and the energy function to optimize the optimized power trading strategy. Neighborhood search to generate a group of adjacent candidate solutions; according to the energy function, calculate the energy values ​​of the current solution and the candidate solutions, if the energy value of the candidate solution is lower, accept the candidate solution as the new current solution; if the energy value of the candidate solution is higher, accept the candidate solution with a certain probability, and the probability decreases as the temperature decreases, to generate an updated current solution; using the updated current solution, gradually reduce the temperature according to the preset cooling rate to help the reinforcement learning model jump out of the local optimal solution, explore a larger solution space, and generate a simulated annealing algorithm state after cooling; when the temperature drops to the termination temperature or reaches a predetermined number of iterations, stop the simulated annealing algorithm, and generate a further optimized power trading strategy based on the simulated annealing algorithm state after cooling to minimize power costs while maintaining high flexibility and adaptability to reduce trading risks.

[0190] In this step, reinforcement learning is a machine learning method that learns how to take actions to maximize a certain cumulative reward through interaction with the environment. In the optimization of electricity trading strategies, the state space of the reinforcement learning model includes information such as the current market electricity price, electricity demand, and the state of the energy storage system; the action space refers to the behavior of deciding when to buy or sell electricity; the reward function is used to evaluate the effect of each decision, such as giving positive or negative feedback based on electricity cost savings. The simulated annealing algorithm is a global optimization technique inspired by the annealing process in metallurgy. It avoids falling into the local optimum by accepting suboptimal solutions, thereby finding a better or even global optimal solution.

[0191] First, based on the existing power dispatching scheme and market electricity price data set, a reinforcement learning model is established. The model will continuously try different power purchase and sale strategies, and adjust the strategy according to the actual effect to minimize the power cost. Then, the trained reinforcement learning model is used to generate a preliminary power trading strategy. Next, the simulated annealing algorithm is used to further optimize this strategy: after initializing the SA parameters, an energy function is defined to evaluate the quality of the strategy, candidate solutions are generated through neighborhood search and accepted by probability, and the temperature is gradually reduced until the termination condition is reached, so as to help jump out of the local optimal solution and find a more robust global optimal solution. Finally, the learning results of the reinforcement learning model and the optimization results of the simulated annealing algorithm are combined to form a set of optimal power trading strategies that comprehensively consider cost savings and risk control.

[0192] In the embodiment of the present application, it is assumed that a certain shelter hospital has the ability to generate electricity independently and can participate in the local electricity market to buy and sell electricity. The hospital management hopes to develop an electricity trading plan that can meet the needs of daily operations and effectively control costs. To this end, they first collected electricity price data for different time periods in the past year to identify price fluctuation patterns. Then, a reinforcement learning model was constructed, in which the state space includes factors such as the current electricity price level, expected electricity consumption, and the remaining capacity of energy storage facilities; the action is the choice of purchasing or selling electricity; and the reward mechanism is set according to the amount of electricity saved. After a period of self-learning, the model learned to increase electricity purchases when electricity prices are low, and reduce or even sell excess electricity during peak hours. In order to further improve the stability of the strategy, the simulated annealing algorithm was introduced to fine-tune the above strategy, and by setting appropriate initial temperature, cooling rate and other parameters, it is ensured that it can respond flexibly even in the face of sudden price changes. The final trading strategy not only significantly reduced the hospital's energy expenditure, but also enhanced its flexibility and risk resistance in participating in the market.

[0193] This application considers that in the power market, optimizing power purchase and sales strategies is crucial to reducing operating costs and improving economic benefits. By introducing market electricity price information and combining reinforcement learning and simulated annealing algorithms, price fluctuation patterns can be automatically learned to generate the optimal power trading strategy. This method not only takes into account the current market conditions, but also adapts to future uncertainties.

[0194] Optionally, the step 103 introduces market electricity price information according to the power dispatching plan, automatically learns the price fluctuation law of the power market through the reinforcement learning algorithm, calculates the power purchase and sale strategy, minimizes the power cost, and further optimizes the decision-making process through the simulated annealing algorithm to reduce the risk of power trading and generate the optimal power trading strategy, including:

[0195] Using the power dispatching scheme, introducing market electricity price information, collecting current and historical market electricity price data, including electricity price fluctuations in different time periods, and electricity price differences between peak and valley periods, to generate a market electricity price data set;

[0196] Based on the market electricity price data set, a reinforcement learning model is established to automatically learn the price fluctuation rules of the electricity market through continuous trial and error, gradually optimize the electricity purchase and sales strategies, and generate a preliminary electricity trading strategy, as follows:

[0197] Initializing a reinforcement learning model, wherein the state space of the reinforcement learning model includes the current market electricity price, electricity demand, and energy storage system status, and the action space includes decisions on electricity purchase and sale;

[0198] Define a reward function to evaluate the performance of the reinforcement learning model in each decision step. The reward function R(t) is:

[0199]

[0200] Among them, C buy (t) is the cost of purchasing a unit of electricity from the external power grid at time t; P buy (t) is the amount of electricity purchased from the external power grid at time t; P RE (t) is the amount of electricity generated by renewable energy at time t; P discharge (t) is the discharge power of the energy storage system at time t; P load (t) is the electricity demand at time t; β is the weight coefficient of the electricity supply and demand balance; γ is the exponential parameter of the supply and demand balance, which is used to adjust the importance of the supply and demand balance; ∈ is a small constant to prevent division by zero errors;

[0201] The reinforcement learning model is trained using the market electricity price data set and the power dispatching plan. The reinforcement learning model will try different strategies for purchasing and selling electricity, provide feedback based on actual results, gradually adjust the strategy to minimize electricity costs, and generate an optimized electricity trading strategy.

[0202] Based on the optimized power trading strategy, the simulated annealing algorithm is used to optimize the decision-making process. By simulating the physical annealing process, the reinforcement learning model is helped to jump out of the local optimal solution, find the global optimal solution, reduce the risk of power trading, and generate a further optimized power trading strategy, as follows:

[0203] Initialize the simulated annealing algorithm and set the initial temperature T 0 , cooling rate α, termination temperature T final ;

[0204] Define the energy function to evaluate the quality of the current electricity trading strategy. The energy function E(t) is:

[0205]

[0206] Among them, δ is the weight coefficient of power supply and demand balance; η is the index parameter of supply and demand balance, which is used to adjust the importance of supply and demand balance;

[0207] Performing a neighborhood search on the optimized power trading strategy to generate a set of neighboring candidate solutions;

[0208] According to the energy function, the energy values ​​of the current solution and the candidate solution are calculated. If the energy value of the candidate solution is lower, the candidate solution is accepted as the new current solution; if the energy value of the candidate solution is higher, the candidate solution is accepted with a certain probability. The probability decreases as the temperature decreases. The specific probability formula is:

[0209]

[0210] Among them, E new is the energy value of the candidate solution; E current is the energy value of the current solution; T is the current temperature; P accept represents the probability of accepting a candidate solution;

[0211] Using the updated current solution, the temperature is gradually reduced according to the preset cooling rate to help the reinforcement learning model jump out of the local optimal solution, explore a larger solution space, and generate the simulated annealing algorithm state after cooling. The temperature update formula is:

[0212] T new =α·T current

[0213] Among them, T new is the new temperature; T current is the current temperature; α is the cooling rate; or the temperature reduction coefficient

[0214] When the temperature drops to the termination temperature or reaches a predetermined number of iterations, the simulated annealing algorithm is stopped, and a further optimized power trading strategy is generated based on the simulated annealing algorithm state after the temperature drops;

[0215] By combining the optimization results of the reinforcement learning model and the further optimization of the simulated annealing algorithm, an optimal electricity trading strategy is generated to minimize electricity costs while maintaining high flexibility and adaptability and reducing trading risks.

[0216] The reward function is used to evaluate the performance of the reinforcement learning model in each decision-making step, encouraging the model to reduce the cost of purchasing electricity and maximize the use of renewable energy and energy storage systems; the energy function is used to evaluate the quality of the current electricity trading strategy, helping the simulated annealing algorithm to jump out of the local optimal solution and find the global optimal solution; the acceptance probability allows the simulated annealing algorithm to accept a worse solution with a certain probability, thereby avoiding falling into the local optimal solution.

[0217] The following is a brief introduction to the design reasons of each sub-item of the formula:

[0218] In the reward function R(t), -C buy (t)·P buy (t) This part represents the cost of purchasing electricity from the external grid. The negative sign means that the higher the cost, the lower the reward; This part shows the ratio of renewable energy generation and energy storage system discharge power to total electricity demand. This ratio is used to measure the balance of electricity supply and demand and encourage the use of more renewable energy and energy storage systems.

[0219] The following is a brief introduction to how to obtain the parameters of the formula:

[0220] Among them, C buy (t) Real-time electricity price data can be obtained from the electricity market or suppliers; buy (t) is the amount of electricity purchased from the external power grid at each time point t, which is a decision variable of the model; β and γ are weight coefficients and exponential parameters, which can be set according to actual needs; P RE (t) and P discharge (t) are the renewable energy generation and energy storage system discharge power, which can be predicted through weather forecasts and historical data; P load (t) is the total electricity demand, which can be obtained through forecasting. ∈ is a small constant to prevent division by zero errors.

[0221] The following is a brief introduction to the design reasons of each sub-item of the formula:

[0222] In the energy function E(t), C buy (t)·P buy (t) This part represents the cost of purchasing electricity from the external grid. In the energy function, the higher the cost, the higher the energy value; This part shows the ratio of renewable energy generation and energy storage system discharge power to total electricity demand. This ratio is used to measure the balance of electricity supply and demand and encourage the use of more renewable energy and energy storage systems.

[0223] The following is a brief introduction to how to obtain the parameters of the formula:

[0224] Among them, δ and η are weight coefficients and exponential parameters, which can be set according to actual needs; C buy (t) Real-time electricity price data can be obtained from the electricity market or suppliers; buy (t) is the amount of electricity purchased from the external power grid at each time point t, which is a decision variable of the model; β and γ are weight coefficients and exponential parameters, which can be set according to actual needs; P RE (t) and P discharge (t) are the renewable energy generation and energy storage system discharge power, which can be predicted through weather forecasts and historical data; P load (t) is the total electricity demand, which can be obtained through prediction.

[0225] The following is a brief introduction to the design reasons of each sub-item of the formula:

[0226] In the acceptance probability formula P accept middle, The reason for designing this sub-item is to calculate the probability of accepting a worse solution based on the current temperature and the energy difference between the new and old solutions, so that the algorithm can accept more worse solutions in the early stage to avoid local optimality, and gradually reduce this probability in the later stage to move towards the global optimality.

[0227] The following is a brief introduction to how to obtain the parameters of this formula:

[0228] Among them, E new and E current are the energy values ​​of the candidate solution and the current solution respectively. T is the current temperature, which gradually decreases with iterations.

[0229] The following is a brief introduction to the design reasons of each sub-item of the formula:

[0230] Update formula at temperature T new In the current The reason for designing this sub-item is to control the balance between exploration and utilization of the algorithm by gradually lowering the temperature, ensuring that the solution space can be extensively explored during the search process and that concentrated optimization can be achieved in the later stage.

[0231] The following is a brief introduction to how to obtain the parameters of the formula:

[0232] Among them, α can be determined through experiments and experience. Generally speaking, the closer the cooling rate is to 1, the slower the temperature drops and the stronger the algorithm's exploration ability is; the smaller the cooling rate is, the faster the temperature drops and the faster the algorithm converges; T current Initial temperature T 0 Usually set to a high value to allow enough room for exploration in the beginning.

[0233] Assume there is a microgrid system with the following parameters:

[0234] Initial temperature T 0 =100; cooling rate α = 0.95; termination temperature T final =0.01; weight coefficient β=10, δ=10; exponential parameter γ=2, η=2; small constant ∈=0.01;

[0235] Table 3 below is the forecast of electricity demand and renewable energy generation:

[0236] Table 3

[0237]

[0238] Table 4 below is the market electricity price information:

[0239] Table 4

[0240]

[0241] The following is the initial energy storage system state: E storage，0 =100kWh

[0242] Table 5 below shows the working modes and power consumption of electrical equipment:

[0243] Table 5

[0244]

[0245] The following is the reinforcement learning model initialization:

[0246] State space: includes current market electricity prices, electricity demand, and energy storage system status.

[0247] Action space: includes decisions on buying and selling electricity.

[0248] Reward function:

[0249]

[0250] The following is the simulated annealing algorithm initialization:

[0251] Initial temperature T 0 =100; cooling rate α = 0.95; termination temperature T final =0.01;

[0252] The following is the simulated annealing algorithm implementation:

[0253] Energy function:

[0254]

[0255] Neighbor-OR search: Generates a set of neighboring candidate solutions.

[0256] Acceptance probability: Calculate the energy values ​​of the current solution and the candidate solution based on the energy function to decide whether to accept the candidate solution.

[0257] Assume that at time t = 0, the following conditions exist:

[0258] P load，0 =100kW; P RE，0 =50kW; P discharge，0 =20kW; C buy (0) = 0.15 yuan / kWh; P buy，0 =30kW;

[0259] The following is the reward function calculation:

[0260]

[0261] R(0)=0.398

[0262] The following is the energy function calculation:

[0263]

[0264] E(0)=-0.398

[0265] The following is the acceptance probability calculation:

[0266] Assume that the energy value of the current solution is E current =-0.398, the energy value of the candidate solution E new =-0.45, current temperature T = 100;

[0267]

[0268] P accept =exp(0.00052)

[0269] P accept ≈1.00052

[0270] At time t = 0, the reward function R(0) = 0.398, indicating that this decision makes good use of renewable energy and energy storage systems while reducing the cost of purchasing electricity. The energy function E(0) = -0.398, indicating that the current solution is a good solution. Acceptance probability P accept ≈1.00052, which is close to 1, meaning that the candidate solution has a high probability of being accepted because its energy value is lower.

[0271] Through the above calculations, we can see that through reinforcement learning and simulated annealing algorithms, we can gradually optimize the electricity purchase and sales strategies, minimize the electricity costs, maintain high flexibility and adaptability, and reduce transaction risks.

[0272] 104. Based on the optimal power trading strategy, the operating status of the power system of the cabin hospital is monitored. When an abnormal situation is detected, the cause of the fault is analyzed and quickly located, the backup power supply is automatically started, and the power distribution plan is adjusted to ensure the continuous power supply of key medical equipment. The abnormal information is recorded to ensure the security and non-tamperability of the data, and the abnormal information is reported to the management platform so that timely response measures can be taken.

[0273] In this step, the power system operation status monitoring is to monitor the various indicators of the power system in real time, such as voltage, current, power, etc. The backup power supply is a backup power supply system used for temporary power supply when the main power supply fails. Data security and non-tamperability ensure that the recorded data is not tampered with and maintain its integrity and authenticity.

[0274] Based on the optimal power trading strategy, the system monitors the operating status of the square cabin hospital power system in real time. Once an abnormal situation is detected, the system will quickly locate the cause of the fault and automatically start the backup power supply. At the same time, the power distribution plan is adjusted to ensure the continuous power supply of key medical equipment. The abnormal information is recorded, and the security and non-tamperability of the data are guaranteed, and the abnormal information is reported to the management platform so that timely countermeasures can be taken.

[0275] In this application example, the power management system of the Fangcang Hospital is equipped with a real-time monitoring module to continuously monitor various indicators of the power system. If a power grid failure is detected, the system will immediately start the diesel generator as a backup power source and adjust the power distribution plan to prioritize the power supply of key medical equipment. All abnormal information will be recorded, and blockchain technology will be used to ensure the security and non-tamperability of the data. Abnormal information will be reported to the management platform in real time, and managers can respond quickly and take appropriate maintenance measures.

[0276] Figure 2 A schematic diagram of the structure of a power coordination management system based on the concept of a virtual power plant in a square cabin hospital is provided for the embodiment of the present application. Figure 2 As shown, the system includes:

[0277] The analysis and evaluation module 21 is used to analyze the historical power consumption patterns and weather data, process the received real-time power demand information of each power-consuming device in the square cabin hospital and the power generation forecast information of the renewable energy power generation unit, improve the accuracy of the power generation forecast, evaluate the uncertainty of power demand, and generate optimized power demand and power generation forecast information;

[0278] The adjustment and optimization module 22 is used to dynamically adjust the working mode of each power-consuming device and the charging and discharging strategy of the energy storage system based on the optimized power demand and power generation forecast information and the current state of the energy storage system in the square cabin hospital by using a mixed integer linear programming model to ensure the balance of power supply and demand and maximize the use of renewable energy, and use a genetic algorithm to optimize the solution process of the mixed integer linear programming model to generate a globally optimal power dispatching plan;

[0279] The calculation generation module 23 is used to introduce market electricity price information according to the power dispatching plan, automatically learn the price fluctuation law of the power market through the reinforcement learning algorithm, calculate the power purchase and sale strategy, minimize the power cost, and further optimize the decision-making process through the simulated annealing algorithm, reduce the risk of power trading, and generate the optimal power trading strategy;

[0280] The monitoring and reporting module 24 is used to monitor the operating status of the power system of the cabin hospital based on the optimal power trading strategy. When an abnormal situation is detected, it analyzes and quickly locates the cause of the fault, automatically starts the backup power supply, and adjusts the power distribution plan to ensure the continuous power supply of key medical equipment, and records the abnormal information to ensure the security and non-tamperability of the data, and reports the abnormal information to the management platform so that timely response measures can be taken.

[0281] Figure 2 The power coordination management system based on the concept of virtual power plant in the square cabin hospital can execute Figure 1 The implementation principle and technical effects of the power coordination management method based on the concept of virtual power plant in the square cabin hospital described in the illustrated embodiment will not be repeated. For the power coordination management system based on the concept of virtual power plant in the square cabin hospital in the above embodiment, the specific way in which each module and unit performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0282] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A power coordination management method based on the concept of virtual power plant in a square cabin hospital, characterized in that: include: Analyze historical electricity consumption patterns and weather data, process the real-time power demand information of each electrical device in the square cabin hospital and the power generation forecast information of renewable energy power generation units, improve the accuracy of power generation forecast, evaluate the uncertainty of power demand, and generate optimized power demand and power generation forecast information; Based on the optimized power demand and power generation forecast information, combined with the current state of the energy storage system in the square cabin hospital, a mixed integer linear programming model is used to dynamically adjust the working mode of each power-consuming equipment and the charging and discharging strategy of the energy storage system to ensure the balance of power supply and demand and maximize the use of renewable energy. A genetic algorithm is used to optimize the solution process of the mixed integer linear programming model to generate a globally optimal power dispatching plan; According to the power dispatching scheme, market electricity price information is introduced, and the price fluctuation law of the power market is automatically learned through the reinforcement learning algorithm, and the power purchase and sale strategies are calculated to minimize the power cost. The decision-making process is further optimized through the simulated annealing algorithm to reduce the risk of power trading and generate the optimal power trading strategy; Based on the optimal power trading strategy, the operating status of the cabin hospital's power system is monitored. When an abnormal situation is detected, the cause of the fault is analyzed and quickly located, the backup power supply is automatically started, and the power distribution plan is adjusted to ensure the continuous power supply of key medical equipment. The abnormal information is recorded to ensure the security and non-tamperability of the data, and the abnormal information is reported to the management platform so that timely response measures can be taken.

2. The electric power coordination management method based on the concept of virtual power plant in the square cabin hospital according to claim 1 is characterized in that: Based on the optimized power demand and power generation forecast information, combined with the current state of the energy storage system in the square cabin hospital, a mixed integer linear programming model is used to dynamically adjust the working mode of each power-consuming device and the charging and discharging strategy of the energy storage system to ensure the balance of power supply and demand and maximize the use of renewable energy, and a genetic algorithm is used to optimize the solution process of the mixed integer linear programming model to generate a globally optimal power dispatching plan, including: Using the optimized power demand and power generation forecast information, preliminarily matching the demand of power-consuming equipment with the power generation capacity of renewable energy sources is performed to obtain a preliminary balance of power supply and demand; Based on the preliminary power supply and demand balance, evaluate the current state of the energy storage system, including parameters of remaining power, maximum charging and discharging rates, and generate an energy storage system state report; Based on the energy storage system status report and the preliminary power supply and demand balance, a mixed integer linear programming model is established with the goal of maximizing the utilization rate of renewable energy and the economy of the power system, wherein the variables include the working mode of the power equipment and the charging and discharging state of the energy storage system, and the constraints include power demand, power generation capacity and energy storage system capacity, to generate a mixed integer linear programming model; Utilizing the mixed integer linear programming model, a genetic algorithm is used to optimize the model solution process to overcome the problem of local optimal solution and generate an optimized solution result; Based on the optimized solution, a globally optimal power dispatching plan is generated to ensure the balance between power supply and demand, while minimizing power costs and improving the flexibility and reliability of the system.

3. The power coordination management method based on the concept of virtual power plant in the square cabin hospital according to claim 2 is characterized in that: According to the energy storage system status report and the preliminary power supply and demand balance, a mixed integer linear programming model is established with the goal of maximizing the utilization rate of renewable energy and the economy of the power system, wherein the variables include the working mode of the power equipment and the charging and discharging state of the energy storage system, and the constraints include power demand, power generation capacity and energy storage system capacity, and the mixed integer linear programming model is generated, including: Using the energy storage system status report and the preliminary balance of power supply and demand, the model objective is determined to maximize the utilization rate of renewable energy and the economy of the power system, that is, to give priority to the use of renewable energy and reduce the cost of purchasing electricity from the external power grid while meeting all power demands; Based on the model objectives, the working mode of the electrical equipment and the charging and discharging state of the energy storage system are defined to generate model variables; According to the model variables, the power demand, power generation capacity and energy storage system capacity are set to generate model constraints; Using the model objectives, model variables and constraints, a mathematical expression is constructed to form a linear equation system; Based on the linear equations, all elements are integrated to generate a complete mixed integer linear programming model as the basis of the optimization algorithm to find the best solution.

4. The electric power coordination management method based on the concept of virtual power plant in the square cabin hospital according to claim 3 is characterized in that: According to the energy storage system status report and the preliminary power supply and demand balance, a mixed integer linear programming model is established with the goal of maximizing the utilization rate of renewable energy and the economy of the power system, wherein the variables include the working mode of the power equipment and the charging and discharging state of the energy storage system, and the constraints include power demand, power generation capacity and energy storage system capacity, and the mixed integer linear programming model is generated, including: Using the energy storage system status report and the preliminary power supply and demand balance, the model objective is determined to maximize the utilization rate of renewable energy and the economy of the power system, that is, on the premise of meeting all power demands, give priority to the use of renewable energy and reduce the cost of purchasing electricity from the external power grid. The objective function minZ is: Among them, C buy is the cost of purchasing a unit of electricity from the external power grid; P buy，t is the amount of electricity purchased from the external power grid at time t; P RE，t is the renewable energy power generation at time t; α is the weight coefficient of renewable energy utilization rate; Based on the model objectives, the working mode of the electrical equipment and the charging and discharging state of the energy storage system are defined to generate model variables; According to the model variables, the power demand, power generation capacity and energy storage system capacity are set to generate the constraints of the model. The following is a set of linear equations of the constraints: Among them, P load，t is the electricity demand at time t; E max is the maximum capacity of the energy storage system; P charge，max is the maximum charging rate of the energy storage system; P discharge，max is the maximum discharge rate of the energy storage system; P i，t is the power consumption of electrical equipment i at time t; η charge is the charging efficiency of the energy storage system; η discharge is the discharge efficiency of the energy storage system; x i，t is the working mode of the electrical equipment i at time t; P charge，t is the charging power of the energy storage system at time t; P discharge，t is the discharge power of the energy storage system at time t; E storage，t is the remaining power of the energy storage system at time t; It means that for all time points t, the relevant equations, inequalities or constraints must be valid at every time point t; Based on the linear equations, all elements are integrated to generate a complete mixed integer linear programming model as the basis of the optimization algorithm to find the best solution.

5. The electric power coordination management method based on the concept of virtual power plant in the square cabin hospital according to claim 2 is characterized in that: The mixed integer linear programming model is used to optimize the solution process of the model using a genetic algorithm to overcome the problem of local optimal solutions and generate optimized solution results, including: Using the mixed integer linear programming model, the population is initialized and a set of initial solutions are randomly generated, each solution represents a possible power dispatching scheme, covering the working mode of the power equipment and the charging and discharging state of the energy storage system; Based on the initial solution, the fitness of the initial solution is evaluated, and the fitness value of the initial solution is calculated to reflect the performance of the initial solution in meeting power demand, maximizing renewable energy utilization, and minimizing power cost; According to the fitness value, some excellent solutions are selected to enter the next generation population, and a roulette wheel selection or a tournament selection method is adopted to obtain the selected solution; Utilizing the selected solutions, pairing and generating new solutions through crossover operations, simulating the gene recombination process in biological genetics, and obtaining a new solution population; Based on the new solution population, a mutation operation is performed on some offspring solutions to randomly change the values ​​of certain variables, increase the diversity of the population, and generate a mutated solution population; Using the mutated solution population, re-performing the fitness evaluation, selection, crossover and mutation operations on the new generation population, and continuously iterating until a stop condition is met, such as reaching a predetermined number of iterations or the fitness value no longer significantly increases; Based on the solutions generated in all iterative processes, the solution with the highest fitness is selected as the global optimal solution to generate the optimized solution result.

6. The electric power coordination management method based on the concept of virtual power plant in the square cabin hospital according to claim 1 is characterized in that: According to the power dispatching scheme, the market electricity price information is introduced, the price fluctuation law of the power market is automatically learned through the reinforcement learning algorithm, the power purchase and sale strategies are calculated to minimize the power cost, and the decision-making process is further optimized through the simulated annealing algorithm to reduce the risk of power trading and generate the optimal power trading strategy, including: Using the power dispatching scheme, introducing market electricity price information, collecting current and historical market electricity price data, including electricity price fluctuations in different time periods, and electricity price differences between peak and valley periods, to generate a market electricity price data set; Based on the market electricity price data set, a reinforcement learning model is established to automatically learn the price fluctuation rules of the electricity market through continuous trial and error, gradually optimize the electricity purchase and sales strategy, and generate a preliminary electricity trading strategy; The reinforcement learning model is trained using the market electricity price data set and the power dispatching plan. The reinforcement learning model will try different strategies for purchasing and selling electricity, provide feedback based on actual results, gradually adjust the strategy to minimize electricity costs, and generate an optimized electricity trading strategy. Based on the optimized power trading strategy, the simulated annealing algorithm is used to optimize the decision-making process. By simulating the physical annealing process, the reinforcement learning model is helped to jump out of the local optimal solution and find the global optimal solution, thereby reducing the risk of power trading and generating a further optimized power trading strategy. The optimization results of the comprehensive reinforcement learning model and the further optimization of the simulated annealing algorithm are used to generate the optimal electricity trading strategy to minimize the electricity cost while maintaining high flexibility and adaptability and reducing trading risks.

7. The method for coordinated power management based on the concept of virtual power plant in a square cabin hospital according to claim 6 is characterized in that: According to the power dispatching scheme, the market electricity price information is introduced, the price fluctuation law of the power market is automatically learned through the reinforcement learning algorithm, the power purchase and sale strategies are calculated to minimize the power cost, and the decision-making process is further optimized through the simulated annealing algorithm to reduce the risk of power trading and generate the optimal power trading strategy, including: Using the power dispatching scheme, introducing market electricity price information, collecting current and historical market electricity price data, including electricity price fluctuations in different time periods, and electricity price differences between peak and valley periods, to generate a market electricity price data set; Based on the market electricity price data set, a reinforcement learning model is established to automatically learn the price fluctuation rules of the electricity market through continuous trial and error, gradually optimize the electricity purchase and sales strategies, and generate a preliminary electricity trading strategy, as follows: Initializing a reinforcement learning model, wherein the state space of the reinforcement learning model includes the current market electricity price, electricity demand, and energy storage system status, and the action space includes decisions on electricity purchase and sale; Define a reward function to evaluate the performance of the reinforcement learning model in each decision step. The reward function R(t) is: Among them, C buy (t) is the cost of purchasing a unit of electricity from the external power grid at time t; P buy (t) is the amount of electricity purchased from the external power grid at time t; P RE (t) is the amount of electricity generated by renewable energy at time t; P discharge (t) is the discharge power of the energy storage system at time t; P load (t) is the electricity demand at time t; β is the weight coefficient of the electricity supply and demand balance; γ is the exponential parameter of the supply and demand balance, which is used to adjust the importance of the supply and demand balance; ∈ is a small constant to prevent division by zero errors; The reinforcement learning model is trained using the market electricity price data set and the power dispatching plan. The reinforcement learning model will try different strategies for purchasing and selling electricity, provide feedback based on actual results, gradually adjust the strategy to minimize electricity costs, and generate an optimized electricity trading strategy. Based on the optimized power trading strategy, the simulated annealing algorithm is used to optimize the decision-making process. By simulating the physical annealing process, the reinforcement learning model is helped to jump out of the local optimal solution, find the global optimal solution, reduce the risk of power trading, and generate a further optimized power trading strategy, as follows: Initialize the simulated annealing algorithm, set the initial temperature T0, cooling rate α, and termination temperature T final ; Define the energy function to evaluate the quality of the current electricity trading strategy. The energy function E(t) is: Among them, δ is the weight coefficient of power supply and demand balance; η is the index parameter of supply and demand balance, which is used to adjust the importance of supply and demand balance; Performing a neighborhood search on the optimized power trading strategy to generate a set of neighboring candidate solutions; According to the energy function, the energy values ​​of the current solution and the candidate solution are calculated. If the energy value of the candidate solution is lower, the candidate solution is accepted as the new current solution; if the energy value of the candidate solution is higher, the candidate solution is accepted with a certain probability. The probability decreases as the temperature decreases. The specific probability formula is: Among them, E new is the energy value of the candidate solution; E current is the energy value of the current solution; T is the current temperature; P accept represents the probability of accepting a candidate solution; Using the updated current solution, the temperature is gradually reduced according to the preset cooling rate to help the reinforcement learning model jump out of the local optimal solution, explore a larger solution space, and generate the simulated annealing algorithm state after cooling. The temperature update formula is: T new =α·T current Among them, T new is the new temperature; T current is the current temperature; α is the cooling rate; or the temperature reduction coefficient. When the temperature drops to the termination temperature or reaches a predetermined number of iterations, the simulated annealing algorithm is stopped, and a further optimized power trading strategy is generated based on the simulated annealing algorithm state after the temperature reduction; By combining the optimization results of the reinforcement learning model and the further optimization of the simulated annealing algorithm, an optimal electricity trading strategy is generated to minimize electricity costs while maintaining high flexibility and adaptability and reducing trading risks.

8. The method for coordinated power management based on the concept of virtual power plant in a square cabin hospital according to claim 6, characterized in that: Based on the market electricity price data set, a reinforcement learning model is established to automatically learn the price fluctuation rules of the electricity market through continuous trial and error, gradually optimize the electricity purchase and sales strategy, and generate a preliminary electricity trading strategy, including: Using the market electricity price data set, a reinforcement learning model is initialized, wherein the state space of the reinforcement learning model includes the current market electricity price, electricity demand, and energy storage system state, and the action space includes decisions on electricity purchase and sale, thereby generating an initialized reinforcement learning model; Based on the market electricity price data set and the power dispatch plan, a reward function is defined to evaluate the performance of the reinforcement learning model in each decision step. Common reward indicators include the level of electricity cost and the degree of balance between electricity supply and demand, and a reward function is generated; The initialized reinforcement learning model is trained using the market electricity price data set, the power dispatching plan and the reward function. The reinforcement learning model will try different strategies for purchasing and selling electricity, provide feedback based on actual results, and gradually adjust the strategy to minimize electricity costs, thereby generating a trained reinforcement learning model. Based on the trained reinforcement learning model, a preliminary electricity trading strategy is generated, which includes electricity purchase and sale decisions in different time periods, aiming to minimize electricity costs and ensure a balance between electricity supply and demand.

9. The method for coordinated power management based on the concept of virtual power plant in a square cabin hospital according to claim 6, characterized in that: Based on the optimized power trading strategy, the simulated annealing algorithm is used to optimize the decision-making process, and by simulating the physical annealing process, the reinforcement learning model is helped to jump out of the local optimal solution, find the global optimal solution, reduce the risk of power trading, and generate a further optimized power trading strategy, including: Using the optimized power trading strategy, the simulated annealing algorithm is initialized, the parameters of the initial temperature, the cooling rate, and the termination temperature are set, and the initialized simulated annealing algorithm is generated; Based on the optimized power trading strategy, an energy function is defined to evaluate the quality of the current power trading strategy. The energy function includes indicators of power cost and the degree of balance between power supply and demand. The lower the energy value, the better the strategy. The energy function is generated. Using the initialized simulated annealing algorithm and the energy function, a neighborhood search is performed on the optimized power trading strategy to generate a set of neighboring candidate solutions; According to the energy function, the energy values ​​of the current solution and the candidate solution are calculated. If the energy value of the candidate solution is lower, the candidate solution is accepted as the new current solution; if the energy value of the candidate solution is higher, the candidate solution is accepted with a certain probability, and the probability decreases as the temperature decreases, to generate an updated current solution; Using the updated current solution, gradually reducing the temperature according to a preset cooling rate, helping the reinforcement learning model to jump out of the local optimal solution, explore a larger solution space, and generate a simulated annealing algorithm state after cooling; When the temperature drops to the termination temperature or reaches a predetermined number of iterations, the simulated annealing algorithm is stopped, and based on the state of the simulated annealing algorithm after the temperature drop, a further optimized electricity trading strategy is generated to minimize electricity costs while maintaining high flexibility and adaptability and reducing trading risks.

10. A power coordination management system based on the concept of virtual power plant in a square cabin hospital, characterized in that: include: The analysis and evaluation module is used to analyze historical power consumption patterns and weather data, process the real-time power demand information of each power-consuming device in the square cabin hospital and the power generation forecast information of the renewable energy power generation unit, improve the accuracy of power generation forecast, evaluate the uncertainty of power demand, and generate optimized power demand and power generation forecast information; An adjustment and optimization module is used to dynamically adjust the working mode of each power-consuming device and the charging and discharging strategy of the energy storage system based on the optimized power demand and power generation forecast information, combined with the current state of the energy storage system in the square cabin hospital, using a mixed integer linear programming model to ensure the balance of power supply and demand and maximize the use of renewable energy, and use a genetic algorithm to optimize the solution process of the mixed integer linear programming model to generate a globally optimal power dispatching plan; A calculation generation module is used to introduce market electricity price information according to the power dispatch plan, automatically learn the price fluctuation law of the power market through the reinforcement learning algorithm, calculate the power purchase and sale strategy, minimize the power cost, and further optimize the decision-making process through the simulated annealing algorithm, reduce the risk of power trading, and generate the optimal power trading strategy; The monitoring and reporting module is used to monitor the operating status of the power system of the cabin hospital based on the optimal power trading strategy. When an abnormal situation is detected, the cause of the fault is quickly located by analysis, the backup power supply is automatically started, and the power distribution plan is adjusted to ensure the continuous power supply of key medical equipment. The abnormal information is recorded to ensure the security and non-tamperability of the data, and the abnormal information is reported to the management platform so that timely response measures can be taken.

Images