Adjustable resource configuration method, system and device and storage medium

Abstract

The invention discloses an adjustable resource configuration method, system and device and a storage medium, and the method comprises the steps: obtaining the operation state parameters of a multi-heterogeneous energy device, and constructing an adjustable resource configuration model based on the operation state parameters; the adjustable resource configuration model takes net income maximization of the multi-energy complementary power generation system as an optimization target; and calculating and solving the adjustable resource configuration model to obtain an optimal configuration decision of the energy storage resources. According to the method, the adjustable resource allocation refined model is constructed according to the characteristics of each power generation and energy storage unit, the operation mode of each power system element in the adjustable resource allocation problem is accurately simulated, the adjustable resource allocation problem is converted into a multi-objective optimization problem solved by using SAC, efficient and accurate solution of the adjustable resource allocation model is realized, and the method is suitable for large-scale popularization and application. The efficiency and precision of resource configuration are improved, and the stability and economy of the system are enhanced.

Description

Technical Field

[0001] This invention relates to the technical field of power system energy storage planning, and in particular to an adjustable resource allocation method, system, device and storage medium. Background Technology

[0002] The integration of a high proportion of renewable energy not only alleviates energy supply pressure but also promotes sustainable local economic development. However, its inherent intermittency and volatility pose unprecedented challenges to the power balance of the power grid. Energy storage systems have gradually gained attention due to their ability to suppress renewable energy fluctuations, improve renewable energy absorption rates, and enhance power system operating efficiency. With the increasing variety of energy storage types, the efficient and accurate solution of adjustable resource allocation models has attracted widespread attention. Traditional optimization algorithms such as dynamic programming, genetic algorithms, and particle swarm optimization can efficiently solve simple models; however, the increasing variety of energy storage types has led to a sharp increase in model complexity, making model solutions more difficult.

[0003] The SAC (Soft Actor-Critic) deep reinforcement learning algorithm, as a trial-and-error-based intelligent algorithm, provides an innovative and efficient solution for solving adjustable resource allocation models. The algorithm learns and dynamically adjusts the action strategy of the energy storage system through real-time interaction between the agent and the power grid environment, utilizing reward signals to quickly obtain the optimal adjustable resource allocation decision. Therefore, this invention proposes an innovative method for solving complex adjustable resource allocation models, thereby improving the efficiency of solving such models. Summary of the Invention

[0004] In view of the aforementioned existing problems, this invention is proposed. Therefore, this invention provides an adjustable resource allocation method, system, device, and storage medium to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] In a first aspect, embodiments of the present invention provide an adjustable resource allocation method, comprising:

[0007] Obtain the operating status parameters of multiple heterogeneous energy devices, and construct an adjustable resource allocation model based on the operating status parameters;

[0008] The adjustable resource allocation model aims to maximize the net revenue of the multi-energy complementary power generation system.

[0009] The optimal allocation decision for energy storage resources is obtained by calculating and solving the adjustable resource allocation model.

[0010] As a preferred embodiment of the adjustable resource allocation method described in this invention, the multi-heterogeneous energy equipment includes: wind farms, photovoltaic power plants, solar thermal power plants, energy storage power plants, as well as electrolyzers and fuel cells.

[0011] As a preferred embodiment of the adjustable resource allocation method of the present invention, the method for obtaining the operating status parameters of multiple heterogeneous energy devices includes: charge-discharge efficiency, capacity range, and upper and lower limits of charge-discharge power; heat-to-electric conversion efficiency, thermal storage capacity range, and upper and lower limits of thermal storage / heat release power; hydrogen production / power generation efficiency, hydrogen storage capacity range, and upper and lower limits of hydrogen production / power generation power.

[0012] As a preferred embodiment of the adjustable resource allocation method described in this invention, the adjustable resource allocation model constraints include: transmission channel capacity constraints, wind power / photovoltaic operation constraints, energy storage power station operation constraints, solar thermal power station and thermal storage system operation constraints, and hydrogen energy system operation constraints.

[0013] As a preferred embodiment of the adjustable resource allocation method described in this invention, the adjustable resource allocation model aims to maximize the net revenue of the multi-energy complementary power generation system, and is expressed as follows:

[0014] maxR = IC cap -C o&m -C dep

[0015] Where R represents the annual net income of the multi-energy complementary power generation system, I represents the annual comprehensive income of the multi-energy complementary power generation system, and C represents the annual net income of the multi-energy complementary power generation system. cap C represents the total investment cost of a multi-energy complementary power generation system under the same annual value. o C represents the annual operation and maintenance cost of a multi-energy complementary power generation system. dep This represents the depreciation cost of a multi-energy complementary power generation system.

[0016] The beneficial effects of this preferred technical solution are that the adjustable resource allocation model takes maximizing the net benefit of the multi-energy complementary power generation system as the optimization objective, and considers multiple constraints such as the capacity constraints of the transmission channel, the operation constraints of wind power / photovoltaic, the operation constraints of the energy storage power station, the operation constraints of the solar thermal power station and thermal storage system, and the operation constraints of the hydrogen energy system. It can achieve precise control of the charging and discharging process of each energy storage unit, optimize the synergy and complementarity between multiple energy storage devices, and thus improve the adjustment flexibility and reliability of the multi-energy complementary power generation system.

[0017] As a preferred embodiment of the adjustable resource allocation method of the present invention, the optimal allocation decision of energy storage resources is obtained by calculating and solving the adjustable resource allocation model. This includes: transforming the adjustable resource allocation problem into a multi-objective optimization problem of the SAC algorithm, and designing the state space, action space and reward function of the SAC algorithm.

[0018] Among them, the state space uses the energy storage capacity of different energy storage devices at the same time period as the key factor to manage energy; the action space selects the charging and discharging power and state parameters of each energy storage device as action values; the reward function is to maximize the annual net income of the multi-energy complementary power generation system.

[0019] Training is performed based on the state space, action space, and reward function. The optimal solution is calculated using the SAC algorithm. The charging and discharging power and state parameters of each energy storage device in each time period are determined by the action values ​​output by the Actor network, thus obtaining the optimal allocation decision of energy storage resources.

[0020] The beneficial effects of this preferred technical solution are that it transforms the complex adjustable resource allocation problem into a multi-objective optimization problem that can be solved by the SAC algorithm, which can achieve effective allocation of energy storage resources, meet the system constraints, maximize net benefits, improve the efficiency and accuracy of resource allocation, and enhance the stability and economy of the system.

[0021] As a preferred embodiment of the adjustable resource allocation method described in this invention, the training based on the state space, action space, and reward function includes:

[0022] Initialize the experience replay pool, Actor network, Critic network, and target Critic network;

[0023] At each time step, the state of the multi-energy complementary power generation system is acquired, an action is selected according to the current Actor policy, the environmental feedback reward is obtained to obtain the next state, and the sample <state, action, reward, next state> quadruple is stored in the experience replay pool. When the experience pool accumulates to the target number of samples, batch training is performed.

[0024] The target batch of samples is randomly drawn from the experience pool for network updates; the Critic network is updated by minimizing the temporal difference error of the Bellman equation; the Actor network is updated by maximizing the weighted sum of expected reward and policy entropy; the target Critic network gradually tracks the Critic network through a soft update method.

[0025] When the change in the value function or strategy function between consecutive iterations is less than a preset threshold or the maximum number of iterations is reached, the iteration ends and the solution of the adjustable resource allocation model is completed.

[0026] Secondly, the present invention provides an adjustable resource allocation system, comprising:

[0027] The model building module is used to obtain the operating status parameters of multiple heterogeneous energy devices and build an adjustable resource allocation model based on the operating status parameters; the adjustable resource allocation model takes maximizing the net revenue of the multi-energy complementary power generation system as the optimization objective.

[0028] The calculation module is used to calculate and solve the adjustable resource configuration model to obtain the optimal configuration decision of energy storage resources.

[0029] Thirdly, the present invention provides an electronic device, comprising:

[0030] Memory and processor;

[0031] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the adjustable resource configuration method.

[0032] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the adjustable resource configuration method.

[0033] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention constructs a refined model of adjustable resource allocation based on the characteristics of each power generation and energy storage unit, accurately simulating the operation mode of each power system component in the adjustable resource allocation problem. It sets up the state space, action space, and reward function of the SAC algorithm. The state space defines all observable states of the multi-energy complementary power generation system, helping the algorithm understand the current environmental situation; the action space determines all possible actions the algorithm can execute, thus affecting its ability to control the environment; the reward function is the standard for evaluating the quality of the actions taken, guiding the algorithm to make better decisions during the learning process. This transforms the adjustable resource allocation problem into a multi-objective optimization problem solved using SAC, thereby achieving efficient and accurate solutions to the adjustable resource allocation model, improving the efficiency and accuracy of resource allocation, and enhancing the stability and economy of the system. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0035] Figure 1This is a schematic flowchart illustrating an adjustable resource configuration method, system, device, and storage medium according to an embodiment of the present invention.

[0036] Figure 2 This is a comparison chart of the reward function values ​​of various algorithms in an adjustable resource allocation method, system, device, and storage medium according to an embodiment of the present invention. Detailed Implementation

[0037] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0038] Example 1, referring to Figure 1 As one embodiment of the present invention, this embodiment provides an adjustable resource configuration method, including:

[0039] S100: Obtain the operating status parameters of multiple heterogeneous energy devices and construct an adjustable resource allocation model based on the operating status parameters; the adjustable resource allocation model takes maximizing the net revenue of the multi-energy complementary power generation system as the optimization objective;

[0040] S200: The adjustable resource allocation model is calculated and solved to obtain the optimal allocation decision of energy storage resources.

[0041] It should be noted that with the increasing variety of energy storage types, the efficient and accurate solution of adjustable resource allocation models has attracted widespread attention. Traditional optimization algorithms such as dynamic programming, genetic algorithms, and particle swarm optimization can efficiently solve simple models. However, the increasing variety of energy storage types leads to a sharp increase in model complexity, making model solving more difficult. This invention constructs a refined adjustable resource allocation model based on the characteristics of each power generation and energy storage unit, accurately simulating the operation of each power system component in the adjustable resource allocation problem. The invention sets up the state space, action space, and reward function of the SAC algorithm. The state space defines all observable states of the multi-energy complementary power generation system, helping the algorithm understand the current environmental situation; the action space determines all possible actions the algorithm can execute, thus affecting its ability to control the environment; the reward function is the standard for evaluating the quality of the actions taken, guiding the algorithm to make better decisions during the learning process. The adjustable resource allocation problem is transformed into a multi-objective optimization problem solved using SAC, thereby achieving an efficient and accurate solution to the adjustable resource allocation model.

[0042] In this embodiment of the application, the heterogeneous energy devices in step S100 include: wind farms, photovoltaic power plants, solar thermal power plants, energy storage power plants, as well as electrolyzers and fuel cells.

[0043] In this embodiment of the application, the operating status parameters of the multiple heterogeneous energy devices obtained in step S100 include: charge and discharge efficiency, capacity range, and upper and lower limits of charge and discharge power; heat-to-electric conversion efficiency, thermal storage capacity range, and upper and lower limits of thermal storage / heat release power; hydrogen production / power generation efficiency, hydrogen storage capacity range, and upper and lower limits of hydrogen production / power generation power.

[0044] It should be noted that, by leveraging the complementary and coordinated relationships in the energy conversion, storage, and release processes between multiple heterogeneous energy devices and energy storage devices, the operating status parameters of the heterogeneous energy devices are obtained. This application comprehensively considers the operating characteristics of multiple heterogeneous energy devices, mainly including the power output characteristics of wind farms and photovoltaic power plants, the thermal energy storage and release characteristics of solar thermal power plants, the charging and discharging characteristics of energy storage power plants, the energy conversion characteristics of hydrogen production by electrolyzers and power generation by fuel cells, as well as the complementary and coordinated energy flow relationships between various devices. Based on the operating characteristics of energy storage devices, solar thermal power plants, and hydrogen storage systems, an adjustable resource allocation model is established, covering key parameters such as the charging and discharging efficiency, capacity range, and upper and lower limits of charging and discharging power of energy storage devices; the heat-to-electricity conversion efficiency, thermal storage capacity range, and upper and lower limits of thermal storage / heat release power of solar thermal power plants; and the hydrogen production / power generation efficiency, hydrogen storage capacity range, and upper and lower limits of hydrogen production / power generation power of hydrogen storage systems.

[0045] In this embodiment of the application, the adjustable resource allocation model constraints in step S100 include: external transmission channel capacity constraints, wind power / photovoltaic operation constraints, energy storage power station operation constraints, solar thermal power station and thermal storage system operation constraints, and hydrogen energy system operation constraints.

[0046] It should be noted that, in order to ensure the safe and stable operation of the multi-energy complementary power generation system, meet the requirements of transmission channel capacity and new energy utilization rate, achieve coordinated balance of the three energy flows of electricity, heat and hydrogen, and improve the overall energy utilization efficiency of the multi-energy complementary system, the constraints considered in the model include the safety constraints of the multi-energy complementary power generation system and the operational constraints of various devices.

[0047] Specifically, the capacity constraint of the external transmission channel is expressed as follows:

[0048]

[0049] Where, p L (t) represents the power generation of the multi-energy complementary power generation system in time period t, p W (t) represents the power generation of the wind farm during time period t, p V (t) represents the power generation of the photovoltaic power station during time period t, p B (t) represents the power generation of the energy storage power station during time period t, pS (t) represents the power generation of the solar thermal power plant during time period t, p H (t) represents the hydrogen production power of the electrolyzer during time period t, p F (t) represents the power generation of the fuel cell during time period t. and These represent the upper and lower limits of the transmission capacity of the multi-energy complementary power generation system.

[0050] Specifically, wind power / solar power operation constraints include power generation range constraints and utilization rate constraints;

[0051] The power generation range constraint is expressed as:

[0052]

[0053] Where, δ W (t) and δ V (t) represents the theoretical normalized output of the wind farm and the photovoltaic power station during time period t, respectively.

[0054] The utilization constraint is expressed as:

[0055]

[0056] Where Θ represents the set of all time periods, which can be set to 8760h in this application; ε represents the maximum curtailment rate of wind power and photovoltaic power.

[0057] Specifically, the operational constraints of energy storage power stations include: charging and discharging power range constraints, charging and discharging state constraints, state of charge constraints, energy storage capacity range constraints, and energy storage duration constraints; among which,

[0058] The charge / discharge power range constraint is expressed as follows:

[0059]

[0060] in, This represents the discharge power of the energy storage power station during time period t. This represents the charging power of the energy storage power station during time period t. This represents the discharge state of the energy storage power station during time period t, and is a 0-1 variable. The charging state of the energy storage power station during time period t is represented by a 0-1 variable, C. B This represents the capacity of the inverter in the energy storage power station, which is a known quantity.

[0061] The charge / discharge state constraints are expressed as follows:

[0062]

[0063] The state of charge constraint is expressed as:

[0064]

[0065] Among them, e B (t) represents the energy storage capacity of the energy storage power station during time period t, which is the optimization variable; and These represent the charging efficiency and discharging efficiency of the energy storage power station, respectively.

[0066] Specifically, taking a tower-type concentrated solar power (CSP) plant as an example, the operational constraints of a CSP plant include collector heat balance constraints, heat storage / heat release power range constraints, heat-to-electricity power conversion constraints, power generation power range constraints, power generation heat balance constraints, operational status logic constraints, heat storage device capacity range constraints, heat storage device heat balance constraints, and heat storage device external transmission capacity range constraints; among these,

[0067] The heat balance constraint of the solar collector is expressed as:

[0068]

[0069] in, This represents the theoretical input power of the solar thermal power plant during time period t. This represents the amount of heat wasted by the solar thermal power plant during time period t. This represents the thermal storage power of the thermal storage device during time period t.

[0070] The power range constraint for heat storage / release is expressed as follows:

[0071]

[0072] in, This represents the heat release power of the thermal storage device during time period t, and is used as an optimization variable; and These represent the maximum heat storage capacity and the maximum heat release capacity of the solar thermal power plant, respectively.

[0073] The thermal-electric power conversion constraint is expressed as:

[0074]

[0075] Where, p S (t) represents the power generation of the solar thermal power plant during time period t, η S This indicates the heat-to-electricity conversion efficiency of a solar thermal power plant.

[0076] The power generation range constraint is expressed as:

[0077]

[0078] Among them, u S (t) represents the operating status of the solar thermal power plant during time period t, and is a 0-1 optimization variable; and These represent the maximum and minimum technical output of the solar thermal power plant, respectively.

[0079] The heat balance constraint for power generation is expressed as:

[0080]

[0081] Where, ν S (t) represents the start-up state of the solar thermal power plant during time period t, and is a 0-1 optimization variable; This indicates the thermal power required to start up a solar thermal power plant.

[0082] The runtime state logical constraints are represented as follows:

[0083]

[0084] The capacity range constraint of the thermal storage device is expressed as follows:

[0085]

[0086] Among them, e S (t) represents the amount of heat stored by the thermal storage device during time period t, and is the optimization variable; E S This indicates the minimum configuration capacity of the thermal storage device.

[0087] The heat balance constraint of the thermal storage device is expressed as:

[0088]

[0089] in, and These represent the heat storage and heat release efficiencies of a solar thermal power plant, respectively. This indicates the power of heat delivered by the thermal storage device to the heat load.

[0090] The constraint on the external transmission capacity of the thermal storage device is expressed as follows:

[0091]

[0092] in, This indicates the upper limit of the thermal energy storage device's external transmission capacity.

[0093] Specifically, the constraints on the operation of hydrogen energy systems include constraints on hydrogen production by electrolyzers, constraints on the power range of electrolyzers, constraints on power generation by fuel cells, constraints on the power range of fuel cells, constraints on the capacity range of hydrogen storage devices, constraints on the capacity balance of hydrogen storage devices, and constraints on the external transmission capacity range of hydrogen storage devices.

[0094] The constraints of hydrogen production in electrolyzers are expressed as follows:

[0095]

[0096] Among them, VH (t) represents the hydrogen production rate of the electrolyzer during time period t, and is the optimization variable; η H This indicates the hydrogen production efficiency of the electrolyzer.

[0097] The power range constraint of the electrolytic cell is expressed as:

[0098]

[0099] Among them, u H (t) represents the operating state of the electrolyzer during time period t, and is a 0-1 optimization variable; and These represent the maximum and minimum technical output of the electrolytic cell, respectively.

[0100] The power generation constraints of fuel cells are expressed as follows:

[0101]

[0102] Among them, V F (t) represents the amount of hydrogen consumed by the fuel cell during time period t, η F This indicates the power generation efficiency of the fuel cell.

[0103] The power range constraint of a fuel cell is expressed as follows:

[0104]

[0105] Where, p F (t) represents the operating state of the fuel cell during time period t, and is a 0-1 optimization variable; and These represent the maximum and minimum technical output of the fuel cell, respectively.

[0106] The capacity range constraint of hydrogen storage devices is expressed as follows:

[0107]

[0108] Among them, e H (t) represents the amount of hydrogen stored in the hydrogen storage device during time period t, E H This indicates the minimum configured capacity of the hydrogen storage device.

[0109] The capacity balance constraint of hydrogen storage devices is expressed as:

[0110]

[0111] in, This indicates the amount of hydrogen transported out by the hydrogen storage device during time period t.

[0112] The constraint on the external transmission capacity of hydrogen storage devices is expressed as follows:

[0113]

[0114] in, This indicates the upper limit of the hydrogen storage device's external transmission capacity.

[0115] In this embodiment of the application, the adjustable resource allocation model in step S100 takes maximizing the net benefit of the multi-energy complementary power generation system as the optimization objective, and is expressed as follows:

[0116] maxR = IC cap -C o&m -C dep

[0117] Where R represents the annual net income of the multi-energy complementary power generation system, I represents the annual comprehensive income of the multi-energy complementary power generation system, and C represents the annual net income of the multi-energy complementary power generation system. cap C represents the total investment cost of a multi-energy complementary power generation system under the same annual value. o C represents the annual operation and maintenance cost of a multi-energy complementary power generation system. dep This represents the depreciation cost of a multi-energy complementary power generation system.

[0118] Furthermore, the annual comprehensive revenue of a multi-energy complementary power generation system includes two parts: revenue from electricity sales and revenue from hydrogen sales, calculated as follows:

[0119]

[0120] Among them, I E and I H λ represents the revenue from electricity sales and hydrogen sales of a multi-energy complementary power generation system, respectively; E and λ H The price of electricity and hydrogen sold by multi-energy complementary power generation systems; p L (t) and The external power and hydrogen volume of the multi-energy complementary power generation system in time period t; t represents the time period number; T represents the number of optimized time periods throughout the year; Δt represents the duration of a unit time period.

[0121] Furthermore, the total investment cost of a multi-energy complementary power generation system under the same annual value is expressed as follows:

[0122]

[0123] Where r represents the discount rate, m represents the operating life of the multi-energy complementary power generation system, and C cap,k Let k = 1, 2, ... 5 represent the investment costs of wind power, photovoltaic, energy storage battery inverters, energy storage battery devices, solar thermal, thermal storage devices, electrolyzers, fuel cells, and hydrogen storage devices, respectively. The calculation method is as follows:

[0124]

[0125] Among them, C W and λW These represent the installed capacity of the wind farm and the unit investment cost, respectively; C V and λ V These represent the installed capacity of a photovoltaic power station and the unit investment cost, respectively; C B and These represent the installed capacity of the energy storage power station and the investment cost per unit installed capacity (inverter capacity), respectively; E B and These represent the battery capacity and unit capacity investment cost of the energy storage power station, respectively; C S and E represents the installed capacity and unit investment cost of a solar thermal power plant, respectively; S and These represent the capacity of the thermal storage device and the investment cost per unit capacity, respectively; C H and These represent the installed capacity of the electrolytic cell and the unit investment cost, respectively; C F and These represent the installed capacity and unit investment cost of fuel cells, respectively; E H and These represent the capacity of the hydrogen storage device and the investment cost per unit capacity, respectively.

[0126] Furthermore, the calculation method for the operation and maintenance costs and depreciation costs of a multi-energy complementary power generation system is expressed as follows:

[0127]

[0128] Where, γ o This represents the operation and maintenance cost coefficient, taken as 2%; γ r This represents the residual value factor, which is set to 10%.

[0129] In this embodiment of the application, the calculation and solution of the adjustable resource allocation model in step S200 to obtain the optimal allocation decision of energy storage resources includes: transforming the adjustable resource allocation problem into a multi-objective optimization problem of the SAC algorithm, and designing the state space, action space and reward function of the SAC algorithm;

[0130] Among them, the state space uses the energy storage capacity of different energy storage devices at the same time period as the key factor to manage energy; the action space selects the charging and discharging power and state parameters of each energy storage device as action values; the reward function is to maximize the annual net income of the multi-energy complementary power generation system.

[0131] Training is performed based on the state space, action space, and reward function. The optimal solution is calculated using the SAC algorithm. The charging and discharging power and state parameters of each energy storage device in each time period are determined by the action values ​​output by the Actor network, thus obtaining the optimal allocation decision of energy storage resources.

[0132] It should be noted that when designing the state space of the adjustable resource allocation model, key factors need to be comprehensively considered to achieve effective energy management, including the energy storage capacity of the energy storage power station in time period t, the heat storage capacity of the thermal storage device in time period t, and the hydrogen storage capacity of the hydrogen storage device in time period t. The state space of the adjustable resource allocation model is represented as follows:

[0133] s t =[e B (t),e S (t),e H (t)]

[0134] Furthermore, to achieve precise control over the charging and discharging processes of each energy storage device, optimize the synergy and complementarity among various energy storage devices, and improve the adjustment flexibility and reliability of the multi-energy complementary power generation system, the charging and discharging power and state parameters of each energy storage unit are selected as action values ​​when designing the action space. These parameters include the power generation of the wind farm in time period t, the power generation of the photovoltaic power station in time period t, the power generation of the solar thermal power station in time period t, the hydrogen production power of the electrolyzer in time period t, the power generation of the fuel cell in time period t, the discharge power of the energy storage power station in time period t, the charging power of the energy storage power station in time period t, the discharge state of the energy storage power station in time period t, the charging state of the energy storage power station in time period t, the energy storage duration of the energy storage power station, and the energy storage state of the solar thermal power station in time period t. The theoretical input power of the solar thermal power plant in time period t, the heat waste power of the thermal storage device in time period t, the heat storage power of the thermal storage device in time period t, the heat release power of the thermal storage device in time period t, the power generation heat output of the solar thermal power plant in time period t, the operating status of the solar thermal power plant in time period t, the start-up status of the thermal power plant in time period t, the heat output power of the thermal storage device to the heat load in time period t, the hydrogen production of the electrolyzer in time period t, the operating status of the electrolyzer in time period t, the hydrogen consumption of the fuel cell in time period t, the operating status of the fuel cell in time period t, and the amount of hydrogen transported by the hydrogen storage device in time period t. The action space of the adjustable resource allocation model is represented as:

[0135]

[0136] Furthermore, the adjustable resource allocation model aims to maximize the net revenue of the multi-energy complementary power generation system under equal annual values. When designing the reward function, two aspects should be considered: first, the negative value of the comprehensive operating cost should be used as a penalty term; second, penalties for violating constraints should be included. The reward function is shown below:

[0137] r = -R-kl

[0138] Where r represents the reward function of the adjustable resource allocation model, k represents the penalty coefficient for violating the constraints, and l represents the number of violations.

[0139] It should be noted that the algorithm is trained based on the state space, action space, and reward function. The optimal solution is calculated using the SAC algorithm. The action values ​​output by the Actor network determine the charging and discharging power and state parameters of each energy storage unit in each time period, thereby achieving optimal allocation decisions for energy storage resources. Through continuous optimization, the algorithm can simultaneously satisfy the constraints of the multi-energy complementary power generation system and maximize the net benefits of the multi-energy complementary power generation system, achieving coordinated and optimal allocation of adjustable resources.

[0140] In this embodiment of the application, step S200, which involves training based on the state space, action space, and reward function, includes:

[0141] Initialize the experience replay pool, Actor network, Critic network, and target Critic network;

[0142] At each time step, the state of the multi-energy complementary power generation system is acquired, an action is selected according to the current Actor policy, the environmental feedback reward is obtained to obtain the next state, and the sample <state, action, reward, next state> quadruple is stored in the experience replay pool. When the experience pool accumulates to the target number of samples, batch training is performed.

[0143] The target batch of samples is randomly drawn from the experience pool for network updates; the Critic network is updated by minimizing the temporal difference error of the Bellman equation; the Actor network is updated by maximizing the weighted sum of expected reward and policy entropy; the target Critic network gradually tracks the Critic network through a soft update method.

[0144] When the change in the value function or strategy function between consecutive iterations is less than a preset threshold or the maximum number of iterations is reached, the iteration ends and the solution of the adjustable resource allocation model is completed.

[0145] In an alternative embodiment, the Actor network is updated by maximizing the weighted sum of the expected reward and the policy entropy; the entropy regularization parameter is adaptively adjusted using gradient descent.

[0146] In another alternative embodiment, the entropy regularization parameter can also be dynamically adjusted based on the learning process performance through a feedback mechanism.

[0147] Specifically, the solution process for the adjustable resource allocation model based on the SAC algorithm is shown in Table 1:

[0148] Table 1. Solution process of the adjustable resource allocation model based on the SAC algorithm

[0149]

[0150]

[0151] It should be noted that the preset threshold, number of iterations, and sample size in the embodiments of this application can be set by comprehensively considering the accuracy requirements and training time.

[0152] In an optional embodiment, when the number of samples accumulates to 1000, 256 samples can be randomly selected from them for network updates;

[0153] In an optional embodiment, the iteration converges when the change of the value function or strategy function between successive iterations is less than 0.01, or when the maximum number of iterations (10,000) is reached, thus completing the solution of the adjustable resource allocation model.

[0154] Example 2: The above example is an illustrative scheme of an adjustable resource allocation method. It should be noted that the technical solution of this adjustable resource allocation system and the technical solution of the above-described adjustable resource allocation method belong to the same concept. Details not described in detail in this example of the adjustable resource allocation system can be found in the description of the above-described adjustable resource allocation method.

[0155] This embodiment provides an adjustable resource allocation system, comprising:

[0156] The model building module is used to obtain the operating status parameters of multiple heterogeneous energy devices and build an adjustable resource allocation model based on the operating status parameters; the adjustable resource allocation model takes maximizing the net revenue of the multi-energy complementary power generation system as the optimization objective.

[0157] The calculation module is used to solve the adjustable resource allocation model to obtain the optimal allocation decision of energy storage resources.

[0158] This embodiment also provides an electronic device applicable to situations involving adjustable resource configuration methods, including:

[0159] The memory and processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the adjustable resource configuration method as proposed in the above embodiments.

[0160] This embodiment also provides a storage medium on which a computer program is stored, which, when executed by a processor, implements the adjustable resource configuration method as proposed in the above embodiments.

[0161] The storage medium proposed in this embodiment and the method for implementing adjustable resource configuration proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0162] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0163] Example 3, referring to Figure 2 This is one embodiment of the present invention. This embodiment highlights the efficiency and accuracy of solving the adjustable resource capacity configuration problem. Through comparative analysis, the beneficial effects of the present invention are verified.

[0164] In this embodiment, the Proximal Policy Optimization (PPO) algorithm and the Deep Deterministic Policy Gradient (DDPG) algorithm are compared and analyzed. The reward function values ​​are as follows: Figure 2 As shown in Table 2, the optimized configuration capacity and benefits are as follows:

[0165] Table 2. Algorithm Configuration Capacity and Revenue

[0166] algorithm SAC PPO DDPG Energy storage / MW·h 100 98 99 Thermal storage / MW·h 1764 1823 1721 Hydrogen storage / t 4.81 4.73 4.94 Net income / thousand yuan 7429 6861 6798

[0167] Reference Figure 2 As can be seen, compared with the PPO and DDPG algorithms, the SAC algorithm has the largest reward function value after converging to the optimal reward function value, proving that the SAC algorithm can obtain the largest comprehensive benefit in the shortest number of iterations while ensuring the safe and stable operation of the power system. As shown in Table 2, the net benefit obtained by solving the adjustable resource allocation model using the SAC algorithm is the largest, proving that its energy storage configuration has the largest comprehensive benefit and is more reasonable than other configuration results.

[0168] This invention constructs a refined model for adjustable resource allocation based on the characteristics of each power generation and energy storage unit, accurately simulating the operation of each power system component in the adjustable resource allocation problem. It sets up the state space, action space, and reward function of the SAC algorithm. The state space defines all observable states of the multi-energy complementary power generation system, helping the algorithm understand the current environmental situation; the action space determines all possible actions the algorithm can execute, thus affecting its ability to control the environment; the reward function is the standard for evaluating the quality of the actions taken, guiding the algorithm to make better decisions during the learning process. This transforms the adjustable resource allocation problem into a multi-objective optimization problem solved using SAC, thereby achieving efficient and accurate solutions to the adjustable resource allocation model, improving the efficiency and accuracy of resource allocation, and enhancing the stability and economy of the system.

[0169] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An adjustable resource allocation method, characterized in that, include: Obtain the operating status parameters of multiple heterogeneous energy devices, and construct an adjustable resource allocation model based on the operating status parameters; The adjustable resource allocation model aims to maximize the net revenue of the multi-energy complementary power generation system. The optimal allocation decision for energy storage resources is obtained by calculating and solving the adjustable resource allocation model.

2. The adjustable resource allocation method as described in claim 1, characterized in that, The heterogeneous energy devices include: wind farms, photovoltaic power plants, solar thermal power plants, energy storage power plants, as well as electrolyzers and fuel cells.

3. The adjustable resource allocation method as described in claim 2, characterized in that, The operational parameters of the multi-heterogeneous energy devices include: charge / discharge efficiency, capacity range, and upper and lower limits of charge / discharge power; heat-to-electric conversion efficiency, thermal storage capacity range, and upper and lower limits of thermal storage / heat release power; hydrogen production / power generation efficiency, hydrogen storage capacity range, and upper and lower limits of hydrogen production / power generation power.

4. The adjustable resource allocation method as described in claim 3, characterized in that, The constraints of the adjustable resource allocation model include: capacity constraints of transmission channels, operation constraints of wind power / photovoltaic power, operation constraints of energy storage power stations, operation constraints of solar thermal power stations and thermal storage systems, and operation constraints of hydrogen energy systems.

5. The adjustable resource allocation method as described in claim 4, characterized in that, The adjustable resource allocation model, with the optimization objective of maximizing the net revenue of the multi-energy complementary power generation system, is expressed as follows: maxR=I-C cap -C o&m -C dep Where R represents the annual net income of the multi-energy complementary power generation system, I represents the annual comprehensive income of the multi-energy complementary power generation system, and C represents the annual net income of the multi-energy complementary power generation system. cap C represents the total investment cost of a multi-energy complementary power generation system under the same annual value. o C represents the annual operation and maintenance cost of a multi-energy complementary power generation system. dep This represents the depreciation cost of a multi-energy complementary power generation system.

6. The adjustable resource allocation method as described in claim 5, characterized in that, The calculation and solution of the adjustable resource allocation model to obtain the optimal allocation decision of energy storage resources includes: transforming the adjustable resource allocation problem into a multi-objective optimization problem of the SAC algorithm, and designing the state space, action space and reward function of the SAC algorithm; Among them, the state space uses the energy storage capacity of different energy storage devices at the same time period as the key factor to manage energy; the action space selects the charging and discharging power and state parameters of each energy storage device as action values; the reward function is to maximize the annual net income of the multi-energy complementary power generation system. Training is performed based on the state space, action space, and reward function. The optimal solution is calculated using the SAC algorithm. The charging and discharging power and state parameters of each energy storage device in each time period are determined by the action values ​​output by the Actor network, thus obtaining the optimal allocation decision of energy storage resources.

7. The adjustable resource allocation method as described in claim 6, characterized in that, The training based on the state space, action space, and reward function includes: Initialize the experience replay pool, Actor network, Critic network, and target Critic network; At each time step, the state of the multi-energy complementary power generation system is acquired, an action is selected according to the current Actor policy, the environmental feedback reward is obtained to obtain the next state, and the sample <state, action, reward, next state> quadruple is stored in the experience replay pool. When the experience pool accumulates to the target number of samples, batch training is performed. The target batch of samples is randomly drawn from the experience pool for network updates; the Critic network is updated by minimizing the temporal difference error of the Bellman equation; the Actor network is updated by maximizing the weighted sum of expected reward and policy entropy; the target Critic network gradually tracks the Critic network through a soft update method. When the change in the value function or strategy function between consecutive iterations is less than a preset threshold or the maximum number of iterations is reached, the iteration ends and the solution of the adjustable resource allocation model is completed.

8. An adjustable resource allocation system, applied to the method described in any one of claims 1-7, characterized in that, include: The model building module is used to obtain the operating status parameters of multiple heterogeneous energy devices and build an adjustable resource configuration model based on the operating status parameters. The adjustable resource allocation model aims to maximize the net revenue of the multi-energy complementary power generation system. The calculation module is used to calculate and solve the adjustable resource configuration model to obtain the optimal configuration decision of energy storage resources.

9. An electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the adjustable resource configuration method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the adjustable resource allocation method according to any one of claims 1 to 7.

Images