Power multiplexing optimization method and device for virtual power plant to participate in multi-element power service, medium and product

Abstract

The invention discloses a power reuse optimization method and device for a virtual power plant to participate in multiple power services, a medium and a product, and relates to the technical field of virtual power plant optimization, and the method comprises the steps: initializing bottom decision variables of an ice storage system in the virtual power plant at all moments in a preset time period; performing loop optimization on the bottom decision variables at each moment in an initialized preset time period by using a base load host power consumption prediction model, a dual-condition host power consumption prediction model, a power multiplexing optimization model of a virtual power plant participating in multi-element power service and optimization model constraints; obtaining a target underlying decision variable of each moment in a preset time period; and performing operation control on the ice storage system in the preset time period by using the target bottom layer decision variables at all moments in the preset time period. According to the invention, the net income of the virtual power plant participating in the multi-element power service is improved.

Description

Technical Field

[0001] This application relates to the technical field of virtual power plant optimization, and particularly to a power multiplexing optimization method, device, medium, and product for a virtual power plant to participate in diversified power services. Background Technique

[0002] With the accelerating global energy transformation, the virtual power plant, as the core carrier of the energy Internet, its ability to participate in diversified power services has become a key technical direction for the construction of smart grids. How to deeply integrate cutting-edge deep learning technologies with power system optimization theories to build a virtual power plant collaborative optimization system adapted to the complex environment of diversified power services is becoming the focus of common concern in the academic and industrial fields.

[0003] As a hub node connecting distributed energy and power services, the optimization decision-making ability of the virtual power plant directly determines the operation efficiency and economic value of the new power system. The current optimization methods for the virtual power plant to participate in transactions such as electricity spot, green electricity, and power auxiliary services mainly include three technical links: resource characteristic modeling, construction of an optimization model for participating in power services, and an optimization model solution method.

[0004] In the resource characteristic modeling link, the mechanism modeling method is generally adopted currently. The operating characteristics of devices such as photovoltaic and energy storage are characterized by mathematical equations. However, due to equipment losses and environmental factor impacts, there will be a phenomenon of cumulative loss distortion, causing the actual operating state of the equipment to deviate from the theoretical derivation results. For resources such as ice storage cooling systems that do not have a strict and accurate mechanism model, the traditional mechanism modeling method is even more inapplicable, and nonlinear or linear fitting methods are commonly used to find the operating characteristic curve, with extremely large errors. Emerging resource modeling methods based on machine learning, such as algorithms like Backpropagation Neural Network (BP) and Long Short-Term Memory (LSTM), have been used for output prediction, but there are defects such as static modeling limitations and multi-source heterogeneous fragmentation. It can be seen that a pure data-driven model is difficult to meet the requirements of virtual power plant resource characteristic modeling.

[0005] In the link of constructing an optimization model for participating in power services, mixed integer programming is mostly used. Although the optimization models for the virtual power plant to participate in a single power grid or auxiliary services have been relatively maturely developed, the power multiplexing optimization method for the virtual power plant to participate in diversified power services has not been considered.

[0006] In the link of optimizing the model solving method, the mainstream methods directly solve using heuristic algorithms or commercial solvers. Heuristic algorithms have defects such as uncontrollable solution quality, weak constraint handling ability, parameter sensitivity, and decision-making black box. When directly solving using commercial solvers, due to the complexity of the model, there are problems such as high difficulty in solving high-dimensional models and low solving efficiency. Particularly prominent is the "fragmentation" problem: on the one hand, the prediction model is separated from the optimization decision-making, resulting in an amplified cumulative error effect; on the other hand, the data-driven method is decoupled from the physical mechanism model, causing a very high risk probability that the decision result violates the power grid security constraints. These defects severely restrict the profit space of virtual power plants in diversified power services.

[0007] In summary, a power reuse optimization method for virtual power plants to participate in diversified power services is needed to solve the problem of reduced profits caused by the deviation between the virtual power plant relying solely on the mechanism model and the actual working conditions. Summary of the Invention

[0008] The purpose of this application is to provide a power reuse optimization method, device, medium, and product for virtual power plants to participate in diversified power services, so as to solve the problem of reduced profits caused by the deviation between the virtual power plant relying solely on the mechanism model and the actual working conditions.

[0009] To achieve the above purpose, the present application provides the following solutions.

[0010] In the first aspect, the present application provides a power reuse optimization method for virtual power plants to participate in diversified power services, including: Initializing the underlying decision variables of the ice storage cooling system in the virtual power plant at each moment during the preset period; Using the base load host power consumption prediction model, dual-mode host power consumption prediction model, power reuse optimization model for virtual power plants to participate in diversified power services, and optimization model constraints, cyclically optimize the underlying decision variables at each moment during the initialized preset period to obtain the target underlying decision variables at each moment during the preset period; Using the target underlying decision variables at each moment during the preset period to control the operation of the ice storage cooling system during the preset period.

[0011] In one embodiment, the ice energy storage system includes: a cold station part, and the cold station part includes: a cooling tower, a cooling water pump, a chilled water pump, a base load host, and a dual-mode host; the underlying decision variables include: the rotational speed ratio of the fan in the cooling tower, the decision variables of the base load host, and the decision variables of the dual-mode host; the decision variables of the base load host include: the chilled water flow rate, the cooling water flow rate, the instantaneous cooling power of the base load host, the cooling water temperature flowing into the base load host, the cooling water temperature flowing out of the base load host, the chilled water temperature flowing into the base load host, and the chilled water temperature flowing out of the base load host; the decision variables of the dual-mode host include: the ethylene glycol flow rate, the cooling water flow rate, the cooling water temperature flowing into the dual-mode host, the cooling water temperature flowing out of the dual-mode host, the ethylene glycol temperature flowing into the dual-mode host, and the ethylene glycol temperature flowing out of the dual-mode host.

[0012] In one embodiment, the process of determining the power consumption prediction model of the base load host includes: Initialize the MHSA-PINN model; the MHSA-PINN model is a model that combines the multi-head self-attention mechanism and the physics-informed neural network; Obtain the first training feature data set, the first training target data set, and the working mechanism constraint set of the base load host; the first training feature data set includes: the actual values of the training feature decision variables of the base load host at multiple first historical moments; the first training target data set includes: the actual values of the power consumption of the base load host at each first historical moment; Use the first training feature data set, the first training target data set, and the working mechanism constraint set of the base load host to train the MHSA-PINN model to obtain the power consumption prediction model of the base load host.

[0013] In one embodiment, the process of determining the power consumption prediction model of the dual-mode host includes: Initialize the MHSA-PINN model; the MHSA-PINN model is a model that combines the multi-head self-attention mechanism and the physics-informed neural network; Obtain the second training feature data set, the second training target data set, and the working mechanism constraint set of the dual-mode host; the second training feature data set includes: the actual values of the training feature decision variables of the dual-mode host at multiple second historical moments; the second training target data set includes: the actual values of the power consumption of the dual-mode host at each second historical moment; Use the second training feature data set, the second training target data set, and the working mechanism constraint set of the dual-mode host to train the MHSA-PINN model to obtain the power consumption prediction model of the dual-mode host.

[0014] In one embodiment, the power multiplexing optimization model includes: ; ; ; ; Among them, is the net income of the virtual power plant; is the total income of the virtual power plant; is the total cost of the virtual power plant; is the total number of moments in the preset time period; is at the moment of , is the electric power of distributed photovoltaics in the virtual power plant for participating in green power trading; is the electricity price of green power trading at the moment of is the time interval; is at the moment of , is the regulated electric power of the ice storage cooling system for participating in peak shaving ancillary services; is the compensation price of peak shaving ancillary services at the moment of is at the moment of , is the electric energy power from the power grid; is the selling electricity price of the power grid at the moment of

[0015] In one embodiment, the optimization model constraints include: ; ; ; ; ; ; ; ; ; ; ; ; ; ; wherein, is the power consumption of the cooling tower; is the speed ratio of the fan in the cooling tower; is the maximum power consumption of the cooling tower during operation; is the power consumption of the cooling water pump; is the density of water; is the acceleration due to gravity; is the cooling water flow rate; is the head of the cooling water pump; is the working efficiency of the cooling water pump; is the motor efficiency; is the frequency converter efficiency; is the power consumption of the chilled water pump; is the chilled water flow rate; is the head of the chilled water pump; is the working efficiency of the chilled water pump; is the ice storage amount in the ice storage tank at time is the ice storage amount in the ice storage tank at time is the heat loss coefficient of the ice storage tank; is the ice storage efficiency; is the ice making cooling power at time is the ice melting cooling power at time is the ice melting efficiency; is the instantaneous heat exchange power of the plate heat exchanger; is the total heat transfer coefficient of the plate heat exchanger; is the heat transfer area of the plate heat exchanger; is the heat transfer temperature difference between the two sides of the fluid in the plate heat exchanger; is the total delay time of chilled water transportation; is the chiller delay time of chilled water transportation; is the riser delay time of chilled water transportation; is the planar delay time of chilled water transportation; is the chiller tube pass of chilled water transportation; is the chiller transportation flow velocity of chilled water; is the riser tube pass of chilled water transportation; is the riser transportation flow velocity of chilled water; is the planar tube pass of chilled water transportation; is the planar transportation flow velocity of chilled water; is the temperature rise of the pump work during chilled water transportation; is the number of series-connected chilled water pumps; is the ratio of the heat generated by the pump to the power consumption; is the specific heat of water; is the power consumption of the central air conditioner in the room; is the outdoor temperature; is the set temperature of the central air conditioner in the room; is the energy efficiency ratio of the central air conditioner; is the equivalent thermal resistance of the central air conditioner; is the power generation of the distributed photovoltaic; is the electric power transmitted from the distributed photovoltaic to the ice storage cooling system; is the power consumption of the ice storage cooling system; is the electric load caused by the terminal temperature control load; is the maximum power generation of the distributed photovoltaic; is the power consumption of the base load host; is the power consumption of the dual-mode host; is the cooling capacity provided by the base load host; is the cooling capacity directly supplied by the dual-mode host for refrigeration; is the cooling capacity loss caused by pipeline transmission; is the cooling capacity required by the terminal temperature control load.

[0016] In one embodiment, the optimization process of the underlying decision variables at each moment in the preset time period at any current cycle number includes: Determine whether the current cycle number is the first time; If the current cycle number is the first time, then: Based on the base load host power consumption prediction model, dual-mode host power consumption prediction model, power multiplexing optimization model, optimization model constraints, and the underlying decision variables at each moment in the initialized preset time period, determine the initial predicted value of the objective function; Determine the initial predicted value of the objective function as the initial optimal value of the objective function; If the current cycle number is not the first time, then: Obtain the underlying decision variables at each moment in the preset time period at the current cycle number. Based on the base load host power consumption prediction model, dual-mode host power consumption prediction model, power multiplexing optimization model, optimization model constraints, and the underlying decision variables at each moment in the preset time period at the current cycle number, determine the predicted value of the objective function at the current cycle number; Determine whether the stop condition is satisfied; the stop condition is that the current cycle number reaches the maximum cycle number or the function difference at the current cycle number is less than the preset value; the function difference at the current cycle number is the difference between the predicted value of the objective function at the current cycle number and the optimal value of the objective function at the previous cycle number; If the stop condition is satisfied, the underlying decision variables at each moment in the preset time period at the current loop count are determined as the target underlying decision variables at each moment in the preset time period.

[0017] In a second aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer program to implement the power multiplexing optimization method for a virtual power plant participating in multiple power services described in any one of the above.

[0018] In a third aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the power multiplexing optimization method for a virtual power plant participating in multiple power services described in any one of the above.

[0019] In a fourth aspect, the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the power multiplexing optimization method for a virtual power plant participating in multiple power services described in any one of the above.

[0020] According to the specific embodiments provided by the present application, the following technical effects are disclosed in the present application: The present application discloses a power multiplexing optimization method, device, medium, and product for a virtual power plant participating in multiple power services. First, the underlying decision variables at each moment in the preset time period of the ice storage cooling system in the virtual power plant are initialized; then, using the base load host power consumption prediction model, the dual-mode host power consumption prediction model, the power multiplexing optimization model for the virtual power plant participating in multiple power services, and the optimization model constraints, the underlying decision variables at each moment in the initialized preset time period are cyclically optimized to obtain the target underlying decision variables at each moment in the preset time period; finally, using the target underlying decision variables at each moment in the preset time period, the operation of the ice storage cooling system in the preset time period is controlled. The present application comprehensively optimizes the underlying decision variables using the base load host power consumption prediction model, the dual-mode host power consumption prediction model, the power multiplexing optimization model for the virtual power plant participating in multiple power services, and the optimization model constraints, solves the problem of reduced profit caused by the deviation between the virtual power plant relying solely on the mechanism model and the actual working conditions, and improves the net income when the virtual power plant participates in multiple power services. Description of the Drawings

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0022] Figure 1 Schematic diagram of the power multiplexing optimization method for a virtual power plant to participate in multiple power services provided by an embodiment of the present application.

[0023] Figure 2 Architecture diagram of the power multiplexing optimization method for a virtual power plant to participate in multiple power services.

[0024] Figure 3 Exemplary application scenario diagram of the power multiplexing optimization method for a virtual power plant to participate in multiple power services.

[0025] Figure 4 Schematic diagram of the ice storage cooling system structure.

[0026] Figure 5 Schematic diagram of the MHSA-PINN model structure.

[0027] Figure 6 Schematic diagram of the structure of a computer device provided by an embodiment of the present application. Detailed implementation manners

[0028] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0029] The purpose of the present application is to provide a power multiplexing optimization method, device, medium and product for a virtual power plant to participate in multiple power services, aiming to solve the problem of reduced profits of the virtual power plant.

[0030] To make the above objects, features and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the drawings and specific implementation manners.

[0031] In an exemplary embodiment, as Figure 1 and Figure 2 shown, a power multiplexing optimization method for a virtual power plant to participate in multiple power services is provided. The power multiplexing optimization method for a virtual power plant to participate in multiple power services is applied to the environment as Figure 3 shown, and the power multiplexing optimization method for a virtual power plant to participate in multiple power services includes Step 1 - Step 3.

[0032] Step 1: Initialize the underlying decision variables of the ice storage cooling system in the virtual power plant at each moment during the preset time period.

[0033] As an optional implementation manner, as Figure 4As shown in the figure, the ice thermal energy storage system includes: a cold station part, which includes: a cooling tower, a cooling water pump, a chilled water pump, a base load chiller and a dual-mode chiller; the bottom-layer decision variables include: the speed ratio of the fan in the cooling tower, the decision variables of the base load chiller and the decision variables of the dual-mode chiller; the decision variables of the base load chiller include: the chilled water flow rate, the cooling water flow rate, the instantaneous cooling power of the base load chiller, the temperature of the cooling water flowing into the base load chiller, the temperature of the cooling water flowing out of the base load chiller, the temperature of the chilled water flowing into the base load chiller and the temperature of the chilled water flowing out of the base load chiller, and the decision variables of the dual-mode chiller include: the ethylene glycol flow rate, the cooling water flow rate, the temperature of the cooling water flowing into the dual-mode chiller, the temperature of the cooling water flowing out of the dual-mode chiller, the temperature of the ethylene glycol flowing into the dual-mode chiller and the temperature of the ethylene glycol flowing out of the dual-mode chiller.

[0034] Specifically, in addition to the cold station part, the ice thermal energy storage system also includes: a pipe network part and a terminal temperature control load part. In addition to the cooling tower, the cooling water pump, the chilled water pump, the base load chiller and the dual-mode chiller, the cold station part also includes: an ice storage tank and a plate heat exchanger, and the cold station part includes three cycles: a cooling water cycle, an ethylene glycol cycle and a chilled water cycle. The cooling tower cools the municipal tap water and the high-temperature cooling water to obtain low-temperature cooling water, and the low-temperature cooling water is transported to the chillers (including the base load chiller and the dual-mode chiller) through the cooling water pump. The chillers absorb the cold energy of the low-temperature cooling water by condensing the refrigerant in the condenser and return the high-temperature cooling water to the cooling tower to form a cooling water cycle. The dual-mode chiller in the chiller releases cold energy by evaporating the refrigerant in the evaporator, cools the high-temperature ethylene glycol to low-temperature ethylene glycol. According to the opening and closing of the ice-making control valve, the low-temperature ethylene glycol flows to the plate heat exchanger for direct cooling of the refrigeration plate heat exchanger or flows to the ice storage tank for ice-making, and finally returns the high-temperature ethylene glycol to the dual-mode chiller to form an ethylene glycol cycle. The base load chiller in the chiller releases cold energy by evaporating the refrigerant in the evaporator. The plate heat exchanger and the ice-melting refrigeration plate heat exchanger absorb the cold energy released by the low-temperature ethylene glycol and cool the high-temperature chilled water to low-temperature chilled water. The chilled water pump transports the low-temperature chilled water through the pipe network to the terminal temperature control load, and after releasing the cold energy, it becomes high-temperature chilled water and returns to the base load chiller and the plate heat exchanger to form a chilled water cycle.

[0035] The pipe network part is composed of conveying pipelines, which transport low-temperature chilled water over a long distance and return high-temperature chilled water.

[0036] The terminal temperature control load part is composed of a central air conditioner. By setting the comfortable temperature according to the user, a temperature adjustable range is formed, which is reflected as a load cooling power adjustable range, and then affects the electric power adjustable capacity of the entire ice thermal energy storage system forward.

[0037] Step 2: Use the base-load host power consumption prediction model, the dual-mode host power consumption prediction model, the power multiplexing optimization model for the virtual power plant to participate in multiple power services, and the optimization model constraints to cyclically optimize the underlying decision variables at each moment in the initialized preset time period, and obtain the target underlying decision variables at each moment in the preset time period.

[0038] As an alternative implementation, the determination process of the base-load host power consumption prediction model includes steps 211 - 213.

[0039] Step 211: Initialize the MHSA-PINN model; the MHSA-PINN model is a model that combines the multi-head self-attention mechanism and the physics-informed neural network.

[0040] Step 212: Obtain the first training feature data set, the first training target data set, and the operating mechanism constraint set of the base-load host; the first training feature data set includes: the actual values of the training feature decision variables of the base-load host at multiple first historical moments; the first training target data set includes: the actual values of the power consumption of the base-load host at each first historical moment.

[0041] Step 213: Use the first training feature data set, the first training target data set, and the operating mechanism constraint set of the base-load host to train the MHSA-PINN model to obtain the base-load host power consumption prediction model.

[0042] Specifically, the actual values of the training feature decision variables of the base-load host at each first historical moment in the first training feature data set are used as the input feature matrix in the form of a matrix and input into the MHSA-PINN model, and the actual values of the power consumption of the base-load host at each first historical moment in the first training target data set are used as the input target matrix in the form of a matrix and input into the MHSA-PINN model.

[0043] As an alternative implementation, the determination process of the dual-mode host power consumption prediction model includes steps 221 - 223.

[0044] Step 221: Initialize the MHSA-PINN model; the MHSA-PINN model is a model that combines the multi-head self-attention mechanism and the physics-informed neural network.

[0045] Step 222: Obtain the second training feature data set, the second training target data set, and the operating mechanism constraint set of the dual-mode host; the second training feature data set includes: the actual values of the training feature decision variables of the dual-mode host at multiple second historical moments; the second training target data set includes: the actual values of the power consumption of the dual-mode host at each second historical moment.

[0046] Specifically, the actual values of the training feature decision variables of the dual-mode main engine at each second historical moment in the second training feature dataset are used as an input feature matrix in the form of a matrix and input into the MHSA-PINN model. The actual values of the power consumption of the dual-mode main engine at each second historical moment in the second training target dataset are used as an input target matrix in the form of a matrix and input into the MHSA-PINN model.

[0047] Step 223: Use the second training feature dataset, the second training target dataset, and the working mechanism constraint set of the dual-mode main engine to train the MHSA-PINN model to obtain a dual-mode main engine power consumption prediction model.

[0048] Specifically, as Figure 5 shown, the MHSA-PINN model includes: an encoder module and a knowledge-data hybrid residual module. The encoder module is implemented based on the multi-head self-attention mechanism (MHSA), and the knowledge-data hybrid residual module is implemented based on the physics-informed neural networks (PINN). The encoder module is responsible for identifying the position feature information of the input feature sequence (corresponding to the decision variables of the base-load main engine in the input base-load main engine power consumption prediction model, corresponding to the decision variables of the dual-mode main engine in the input dual-mode main engine power consumption prediction model) and outputting the target fitting sequence (corresponding to the predicted value of the electric power of the base-load main engine in the input base-load main engine power consumption prediction model). The knowledge-data hybrid residual module is responsible for comparing the target fitting sequence output by the encoder module with the real target sequence (corresponding to the actual value of the electric power of the base-load main engine in the input base-load main engine power consumption prediction model, corresponding to the actual value of the electric power of the dual-mode main engine in the input dual-mode main engine power consumption prediction model), and at the same time collaborating with the mechanism model (corresponding to the working mechanism constraint set of the base-load main engine in the base-load main engine power consumption prediction model, corresponding to the working mechanism constraint set of the dual-mode main engine in the dual-mode main engine power consumption prediction model) to consider the fitting accuracy, and correcting the internal hyperparameters of the encoder module according to the accuracy feedback to improve the fitting accuracy.

[0049] Generally, the process of the encoder module calculating and capturing the correlation of each feature and each position for the input data is as follows: (1) (2) (3) (4) (5) (6) (7) Among them, is the query matrix; is the input feature matrix, a matrix composed of input feature sequences; is the time step length of the input feature sequence; is the number of features of the input feature sequence; is the weight matrix of the query; is the key matrix; is the weight matrix of the key; is the value matrix; is the weight matrix of the value; is the attention weight matrix; is the softmax function; is the transpose; is the dimension of the query matrix; is the output of the self-attention mechanism; is the output of the th head; is the query matrix of the th head; is the key matrix of the th head; is the value matrix of the th head; is the concatenation operation; is the output of the first head; is the output of the second head; is the weight matrix.

[0050] For the self-attention mechanism, the matrix first generates the query matrix , the key matrix and the value matrix through linear transformation. Then, the similarity between the query and the key is calculated using the dot product, and the attention weight matrix is generated. The output of the self-attention mechanism is obtained by multiplying the attention weight matrix with the value matrix . The output contains the relationship information between different positions in a single input feature sequence; for the multi-head self-attention mechanism, the matrix is decomposed into multiple heads. Each head independently calculates the self-attention. The outputs of all heads are then concatenated and linearly transformed to generate the final output of the multi-head self-attention mechanism. This output contains the relationship information between different positions in all input feature sequences.

[0051] In the knowledge-data hybrid residual module, the calculation formula for the knowledge-data hybrid residual is as follows: (8) (9) where is the knowledge-data hybrid residual; is the distribution weight coefficient of the data residual; is the data residual; is the distribution weight coefficient of the knowledge residual; is the knowledge residual; is the input target matrix, i.e., the true target sequence; is the target fitting sequence.

[0052] For the data residual, its goal is to minimize the deviation between the fitting target matrix generated by the MHSA-PINN model and the input target matrix ; for the knowledge residual, its goal is to enable the MHSA-PINN model to reasonably configure weight parameters based on the mechanism model, thereby reducing the influence of input features that violate physical principles and improving the accuracy of expression fitting. In the context of fitting the base-load host power consumption prediction model and the dual-condition host power consumption prediction model, the physical constraints are represented by energy conservation. The knowledge residual part in the knowledge-data hybrid residual module of the base-load host power consumption prediction model is filled with the working mechanism constraint set of the base-load host, and the knowledge residual part in the knowledge-data hybrid residual module of the dual-condition host power consumption prediction model is filled with the working mechanism constraint set of the dual-condition host. The working mechanism constraint set of the base-load host is shown in Equation (10), and the working mechanism constraint set of the dual-condition host is shown in Equation (11).

[0053] (10) (11) where is the cooling capacity exchanged by the cooling water through the base-load host; is the specific heat of water; is the mass of the cooling water; is the temperature of the cooling water flowing out of the base-load host; is the temperature of the cooling water flowing into the base-load host; is the cooling capacity exchanged by the chilled water through the base-load host; is the mass of the chilled water; is the temperature of the chilled water flowing into the base-load host; is the temperature of the chilled water flowing out of the base-load host; is the power consumption of the base-load host; is the actual refrigeration coefficient of the base-load host; is the minimum refrigeration coefficient of the base load host; is the maximum refrigeration coefficient of the base load host; is the cooling capacity exchanged by the cooling water through the dual-mode host; is the temperature of the cooling water flowing out of the dual-mode host; is the temperature of the cooling water flowing into the dual-mode host; is the cooling capacity exchanged by the ethylene glycol through the dual-mode host; is the specific heat of the ethylene glycol; is the mass of the ethylene glycol; is the temperature of the ethylene glycol flowing into the dual-mode host; is the temperature of the ethylene glycol flowing out of the dual-mode host; is the power consumption of the dual-mode host; is the actual refrigeration coefficient of the dual-mode host; is the minimum refrigeration coefficient of the dual-mode host; is the maximum refrigeration coefficient of the dual-mode host.

[0054] The expression of the power consumption prediction model of the base load host is: (12) Wherein, is the predicted value of the power consumption of the base load host; is the mapping function of the power consumption prediction model of the base load host; is the instantaneous cooling power of the base load host; is the cooling water flow rate; is the chilled water flow rate.

[0055] The expression of the power consumption prediction model of the dual-mode host is: (13) Wherein, is the predicted value of the power consumption of the dual-mode host; is the mapping function of the power consumption prediction model of the dual-mode host; is the instantaneous cooling power of the dual-mode host.

[0056] Step 213 specifically includes: The encoder module first disassembles the first training feature dataset into multiple single-feature sequences and maps them to different subspaces to capture the dependencies at different positions among the feature sequences. Through the feed-forward fully connected layer, it maps out the predicted value of the power consumption of the base-load host. The knowledge-data hybrid residual module compares the predicted value of the power consumption of the base-load host with the training target dataset, calculates the mean squared error to obtain the data residual; the knowledge-data hybrid residual module calculates the knowledge residual according to the operating mechanism constraint set of the base-load host, the first training feature dataset, and the first training target dataset; the knowledge-data hybrid residual module performs weighted summation of the data residual and the knowledge residual to obtain the knowledge-data hybrid residual, and returns it to the encoder module to update the parameters; the MHSA-PINN model is trained in multiple loops, and finally the power consumption prediction model of the base-load host is obtained and saved.

[0057] Step 223 specifically includes: The encoder module first disassembles the second training feature dataset into multiple single-feature sequences and maps them to different subspaces to capture the dependencies at different positions among the feature sequences. Through the feed-forward fully connected layer, it maps out the predicted value of the power consumption of the dual-mode host. The knowledge-data hybrid residual module compares the predicted value of the power consumption of the dual-mode host with the training target dataset, calculates the mean squared error to obtain the data residual; the knowledge-data hybrid residual module calculates the knowledge residual according to the operating mechanism constraint set of the dual-mode host, the second training feature dataset, and the second training target dataset; the knowledge-data hybrid residual module performs weighted summation of the data residual and the knowledge residual to obtain the knowledge-data hybrid residual, and returns it to the encoder module to update the parameters; the MHSA-PINN model is trained in multiple loops, and finally the power consumption prediction model of the dual-mode host is obtained and saved.

[0058] As an optional implementation, the power reuse optimization model includes: (14) (15) (16) (17) Among them, is the objective function; is the net profit of the virtual power plant; is the total profit of the virtual power plant; is the total cost of the virtual power plant; is the total number of time points in the preset time period; is at time point , is the electric power of the distributed photovoltaics in the virtual power plant used to participate in green power trading; is The electricity price of green power trading at a certain moment; is the time interval; is at a certain moment , is the regulated electric power of the ice storage cooling system participating in peak shaving ancillary services; is the compensation price of peak shaving ancillary services at a certain moment; is at a certain moment , is the electric energy power from the power grid; is the selling electricity price of the power grid at a certain moment.

[0059] Specifically, the objective function is to maximize the net income of the virtual power plant participating in multiple power services.

[0060] As an optional implementation method, the optimization model constraints include: (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) Among them, is the power consumption of the cooling tower; is the speed ratio of the fan in the cooling tower; is the maximum power consumption of the cooling tower during operation; is the power consumption of the cooling water pump; is the density of water; is the acceleration due to gravity; is the cooling water flow rate; is the head of the cooling water pump; is the working efficiency of the cooling water pump; is the motor efficiency; is the frequency converter efficiency; is the power consumption of the chilled water pump; is the chilled water flow rate; is the head of the chilled water pump; is the working efficiency of the chilled water pump; is the ice storage capacity of the ice storage tank at time is the ice storage capacity of the ice storage tank at time is the heat loss coefficient of the ice storage tank; is the ice storage efficiency; is the ice making cooling power at time is the ice melting cooling power at time is the ice melting efficiency; is the instantaneous heat exchange power of the plate heat exchanger; is the total heat transfer coefficient of the plate heat exchanger; is the heat transfer area of the plate heat exchanger; is the heat transfer temperature difference between the two sides of the fluid of the plate heat exchanger; is the total delay of chilled water transportation; is the chiller delay of chilled water transportation; is the riser delay of chilled water transportation; is the planar delay of chilled water transportation; is the chiller tube pass of chilled water transportation; is the chiller transportation flow velocity of chilled water; is the riser tube pass of chilled water transportation; is the riser transportation flow velocity of chilled water; is the planar tube pass of chilled water transportation; is the planar transportation flow velocity of chilled water; is the temperature rise of the water pump during chilled water transportation; is the number of series-connected chilled water pumps; is the proportion of heat generated by the water pump in the power consumption; is the specific heat of water; is for the room the power consumption of the central air conditioner in; is the outdoor temperature; is for the room the set temperature of the central air conditioner in; is the energy efficiency ratio of the central air conditioner; is the equivalent thermal resistance of the central air conditioner; is the power generation of the distributed photovoltaic; is the electric power delivered by distributed photovoltaics to the ice thermal energy storage system; is the power consumption of the ice thermal energy storage system; is the electrical load caused by the terminal temperature-controlled load; is the maximum power generation of distributed photovoltaics; is the power consumption of the base load host; is the power consumption of the dual-mode host; is the cooling capacity provided by the base load host; is the cooling capacity directly supplied by the dual-mode host for refrigeration; is the cooling capacity loss caused by pipeline transmission; is the cooling capacity required by the terminal temperature-controlled load.

[0061] Specifically, Equation (18) represents the working model of the cooling tower, Equation (19) represents the working model of the cooling water pump, Equation (20) represents the working model of the chilled water pump, Equation (21) represents the working model of the ice storage tank, Equation (22) represents the working model of the plate heat exchanger, the pipe network model consists of the pipe network time delay and temperature rise represented by Equations (23) and (24), and the terminal temperature-controlled load model is represented by Equation (25). Equation (26) is the complementary and collaborative constraint between distributed photovoltaics and the ice thermal energy storage system, Equation (27) is the power reuse constraint, and Equations (26) and (27) together represent that distributed photovoltaic power generation can supply electricity to the ice thermal energy storage system and participate in green electricity trading for power sales; Equations (28), (29), and (31) are the power and energy interaction constraints between the virtual power plant and multiple markets, indicating that the ice thermal energy storage system needs to purchase electricity from the power grid to meet the load demand, and the ice thermal energy storage system participates in peak shaving auxiliary services through ice storage and ice melting actions; Equation (30) is the total power consumption constraint of the ice thermal energy storage system.

[0062] Specifically, the electrical energy produced by distributed photovoltaics participates in peak shaving auxiliary services by being delivered to the ice thermal energy storage system; at the same time, some of the green electricity produced participates in green electricity trading; realizing the collaborative complementarity and power reuse between distributed photovoltaics and the ice thermal energy storage system.

[0063] As an alternative implementation, the optimization process of the underlying decision variables at each moment in the preset time period of any current cycle number includes: Determine whether the current cycle number is the first time.

[0064] If the current cycle number is the first time, then: Based on the base load host power consumption prediction model, dual-mode host power consumption prediction model, power reuse optimization model, optimization model constraints, and the initialized underlying decision variables at each moment in the preset time period, determine the initial predicted value of the objective function.

[0065] Determine the initial predicted value of the objective function as the initial optimal value of the objective function.

[0066] If the current loop count is not the first time, then: Obtain the underlying decision variables at each moment in the preset time period under the current loop count. Based on the base-load host power consumption prediction model, the dual-condition host power consumption prediction model, the power reuse optimization model, the optimization model constraints, and the underlying decision variables at each moment in the preset time period under the current loop count, determine the predicted value of the objective function under the current loop count.

[0067] Judge whether the stop condition is satisfied; the stop condition is that the current loop count reaches the maximum loop count or the function difference under the current loop count is less than the preset value; the function difference under the current loop count is the difference between the predicted value of the objective function under the current loop count and the optimal value of the objective function under the previous loop count.

[0068] If the stop condition is satisfied, then determine the underlying decision variables at each moment in the preset time period under the current loop count as the target underlying decision variables at each moment in the preset time period.

[0069] Step 3: Use the target underlying decision variables at each moment in the preset time period to control the operation of the ice storage cooling system in the preset time period.

[0070] As can be seen from the above embodiments, the advantages of the present application are as follows.

[0071] (1) The refined modeling method of the ice storage cooling system in the virtual power plant based on MHSA-PINN of the present application. For the component structure with a clear and perfect mechanism model, a traditional mechanism model is established; for the component structure without a clear and perfect mechanism model (i.e., the base-load host and the dual-condition host), an MHSA-PINN is invented. The encoder module based on the multi-head self-attention mechanism inside the MHSA-PINN captures the feature relationships of the input historical data, searches for a reasonable and optimal relationship model, breaks through the shackles of the traditional fuzzy mechanism model, and searches for a relationship model with higher accuracy; the knowledge-data hybrid residual module inside the MHSA-PINN evaluates and optimizes the model output by the encoder, and feeds back to the encoder module based on the multi-head self-attention mechanism to optimize the internal hyperparameters, so that the MHAS-PINN can output a better model. The knowledge residual represents the compliance degree of the output model with the traditional mechanism constraints, ensuring that the output model does not violate the most basic working mechanism constraints. The data residual represents the accuracy of the output model's resource expression. The knowledge residual and the data residual are weighted and summed to form the knowledge-data hybrid residual. By reasonably allocating weights, the degree of attention of the relationship model output by the MHSA-PINN to the working mechanism and accuracy can be controlled. Generally speaking, the refined modeling method of the ice storage cooling system inside the virtual power plant based on the deep learning MHSA-PINN breaks through the shackles of the traditional model and is more concise and accurate while ensuring compliance with the basic working mechanism constraint set.

[0072] (2) The power multiplexing optimization method for the virtual power plant of this application to participate in diversified power services first establishes the constraints on the complementary and collaborative relationship of various resources within the virtual power plant and the power multiplexing constraints, enabling the flow and complementarity of various energies of each resource, and allowing each resource and energy to participate in diversified power services simultaneously or gradually, improving the energy utilization efficiency. Then, it establishes the energy interaction constraints for the virtual power plant to participate in diversified power services, enabling the virtual power plant to simultaneously participate in various electricity energy transactions such as green power trading and peak shaving auxiliary services. Supplementary to the complementary collaboration and power multiplexing of the internal resources of the virtual power plant, it can maximize the reduction of the operation and maintenance costs of the virtual power plant, increase the profit of the virtual power plant participating in diversified power services, reduce carbon emissions, and protect the environment.

[0073] (3) When optimizing the underlying decision variables in this application, the MHSA-PINN model is called multiple times during the optimization solution process. Compared with the traditional optimization solution method, the MHSA-PINN model is embedded in the iterative optimization process, streamlining the solution steps, improving the solution efficiency, and making the optimized target underlying decision variables more accurate.

[0074] In an exemplary embodiment, a computer device is provided, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer program to implement the power multiplexing optimization method for the virtual power plant to participate in diversified power services.

[0075] In an exemplary embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, it implements the power multiplexing optimization method for the virtual power plant to participate in diversified power services.

[0076] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, it implements the power multiplexing optimization method for the virtual power plant to participate in diversified power services.

[0077] In an exemplary embodiment, a computer device is provided. This computer device can be a server or a terminal, and its internal structure diagram can be as Figure 6As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a power multiplexing optimization method for a virtual power plant to participate in multiple power services.

[0078] Those skilled in the art can understand that Figure 6 the structure shown in is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0079] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memory can include Read-Only Memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0080] The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0081] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0082] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope described in this specification.

[0083] In this text, specific examples are used to elaborate on the principles and implementation manners of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application. At the same time, for those of ordinary skill in the art, according to the idea of the present application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A power multiplexing optimization method for a virtual power plant to participate in diversified power services, characterized in that The power reuse optimization method for the virtual power plant to participate in diversified power services includes: Initializing the underlying decision variables of the ice storage cooling system in the virtual power plant at each moment during the preset period; Using the base load host power consumption prediction model, the dual-mode host power consumption prediction model, the power reuse optimization model for the virtual power plant to participate in diversified power services, and the constraints of the optimization model, cyclically optimizing the underlying decision variables at each moment during the initialized preset period to obtain the target underlying decision variables at each moment during the preset period; Using the target underlying decision variables at each moment during the preset period to control the operation of the ice storage cooling system during the preset period.

2. The power multiplexing optimization method for a virtual power plant to participate in diversified power services according to claim 1, wherein The ice storage cooling system includes: a cold station part, and the cold station part includes: a cooling tower, a cooling water pump, a chilled water pump, a base load host, and a dual-mode host; the underlying decision variables include: the speed ratio of the fan in the cooling tower, the decision variables of the base load host, and the decision variables of the dual-mode host; the decision variables of the base load host include: the chilled water flow rate, the cooling water flow rate, the instantaneous cooling power of the base load host, the cooling water temperature flowing into the base load host, the cooling water temperature flowing out of the base load host, the chilled water temperature flowing into the base load host, and the chilled water temperature flowing out of the base load host; the decision variables of the dual-mode host include: the ethylene glycol flow rate, the cooling water flow rate, the cooling water temperature flowing into the dual-mode host, the cooling water temperature flowing out of the dual-mode host, the ethylene glycol temperature flowing into the dual-mode host, and the ethylene glycol temperature flowing out of the dual-mode host.

3. The power multiplexing optimization method for a virtual power plant to participate in diversified power services according to claim 2, wherein The determination process of the base load host power consumption prediction model includes: Initializing the MHSA-PINN model; the MHSA-PINN model is a model combining the multi-head self-attention mechanism and the physics-informed neural network; Obtaining the first training feature data set, the first training target data set, and the working mechanism constraint set of the base load host; the first training feature data set includes: the actual values of the training feature decision variables of the base load host at multiple first historical moments; the first training target data set includes: the actual values of the power consumption of the base load host at each first historical moment; Using the first training feature data set, the first training target data set, and the working mechanism constraint set of the base load host to train the MHSA-PINN model to obtain the base load host power consumption prediction model.

4. The power multiplexing optimization method for a virtual power plant to participate in diversified power services according to claim 2, wherein The determination process of the dual-mode host power consumption prediction model includes: Initializing the MHSA-PINN model; the MHSA-PINN model is a model combining the multi-head self-attention mechanism and the physics-informed neural network; Obtaining the second training feature data set, the second training target data set, and the working mechanism constraint set of the dual-mode host; the second training feature data set includes: the actual values of the training feature decision variables of the dual-mode host at multiple second historical moments; the second training target data set includes: the actual values of the power consumption of the dual-mode host at each second historical moment; Using the second training feature data set, the second training target data set, and the working mechanism constraint set of the dual-mode host to train the MHSA-PINN model to obtain the dual-mode host power consumption prediction model.

5. The power multiplexing optimization method for a virtual power plant to participate in diversified power services according to claim 2, wherein, The power reuse optimization model includes: ； ； ； ； Among them, is the objective function; is the net income of the virtual power plant; is the total income of the virtual power plant; is the total cost of the virtual power plant; is the total number of moments in the preset time period; is at the moment of , is the electric power of distributed photovoltaics in the virtual power plant for participating in green power trading; is the electricity price of green power trading at the moment of is the time interval; is at the moment of , is the regulating electric power of the ice storage cooling system for participating in peak shaving ancillary services; is the compensation price of peak shaving ancillary services at the moment of is at the moment of , is the electric energy power from the power grid; is the power grid's electricity selling price at the moment of 6. The power reuse optimization method for a virtual power plant to participate in diversified power services according to claim 5, wherein The optimization model constraints include: ； ； ； ； ； ； ； ； ； ； ； ； ； ； Among them, is the power consumption of the cooling tower; is the speed ratio of the fan in the cooling tower; is the maximum power consumption of the cooling tower during operation; is the power consumption of the cooling water pump; is the density of water; is the acceleration due to gravity; is the cooling water flow rate; is the head of the cooling water pump; is the working efficiency of the cooling water pump; is the motor efficiency; is the frequency converter efficiency; is the power consumption of the chilled water pump; is the chilled water flow rate; is the head of the chilled water pump; is the working efficiency of the chilled water pump; is the ice storage amount of the ice storage tank at time is the ice storage amount of the ice storage tank at time is the heat loss coefficient of the ice storage tank; is the ice storage efficiency; is the ice making cooling power at time is the ice melting cooling power at time is the ice melting efficiency; is the instantaneous heat exchange power of the plate heat exchanger; is the total heat transfer coefficient of the plate heat exchanger; is the heat transfer area of the plate heat exchanger; is the heat transfer temperature difference between the two sides of the fluid of the plate heat exchanger; is the total delay of chilled water transportation; is the chiller delay of chilled water transportation; is the riser delay of chilled water transportation; is the planar delay of chilled water transportation; is the chiller tube pass of chilled water transportation; is the chiller transportation flow velocity of chilled water; is the riser tube pass of chilled water transportation; is the riser transportation flow velocity of chilled water; is the planar tube pass of chilled water transportation; is the planar transportation flow velocity of chilled water; is the temperature rise of the water pump during chilled water transportation; is the number of series-connected chilled water pumps; is the proportion of heat generated by the pump in the power consumption; is the specific heat of water; is the power consumption of the central air conditioner in the room ; is the outdoor temperature; is the set temperature of the central air conditioner in the room ; is the energy efficiency ratio of the central air conditioner; is the equivalent thermal resistance of the central air conditioner; is the power generation power of the distributed photovoltaic; is the electric power transmitted from the distributed photovoltaic to the ice storage cooling system; is the power consumption of the ice storage cooling system; is the electric load caused by the terminal temperature control load; is the maximum power generation power of the distributed photovoltaic; is the power consumption of the base load host; is the power consumption of the dual-mode host; is the cooling capacity provided by the base load host; is the cooling capacity directly supplied by the dual-mode host for refrigeration; is the cooling capacity loss caused by pipeline transmission; is the cooling capacity required by the terminal temperature control load.

7. The power reuse optimization method for a virtual power plant to participate in diversified power services according to claim 6, characterized in that, The optimization process of the underlying decision variables at each moment in a preset time period at any current loop count includes: Determine whether the current loop count is the first time; If the current loop count is the first time, then: Based on the base-load host power consumption prediction model, dual-mode host power consumption prediction model, power reuse optimization model, optimization model constraints, and the initialized underlying decision variables at each moment in the preset time period, determine the initial predicted value of the objective function; Determine the initial predicted value of the objective function as the initial optimal value of the objective function; If the current loop count is not the first time, then: Obtain the underlying decision variables at each moment in the preset time period at the current loop count. Based on the base-load host power consumption prediction model, dual-mode host power consumption prediction model, power reuse optimization model, optimization model constraints, and the underlying decision variables at each moment in the preset time period at the current loop count, determine the predicted value of the objective function at the current loop count; Determine whether the stop condition is met; the stop condition is that the current loop count reaches the maximum loop count or the function difference at the current loop count is less than a preset value; the function difference at the current loop count is the difference between the predicted value of the objective function at the current loop count and the optimal value of the objective function at the previous loop count; If the stop condition is met, then determine the underlying decision variables at each moment in the preset time period at the current loop count as the target underlying decision variables at each moment in the preset time period.

8. A computer device, comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the computer program to implement the power reuse optimization method for a virtual power plant participating in multiple power services according to any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the power reuse optimization method for a virtual power plant participating in multiple power services according to any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the power reuse optimization method for a virtual power plant participating in multiple power services according to any one of claims 1-7.

Images