Virtual power plant control method, system and device based on energy cluster and energy storage cooperation

Abstract

The invention relates to the technical field of virtual power plant control, and provides a virtual power plant control method, system and device based on energy cluster and energy storage cooperation. The method comprises the following steps: collecting real-time data of energy equipment, and preprocessing the real-time data to obtain preprocessed real-time data; performing prediction by using an LSTM attention map neural network according to the preprocessed real-time data to obtain a prediction result; performing multi-element energy storage system configuration and collaborative optimization according to the prediction result to obtain a charging and discharging power reference value; and adjusting the generator set in the virtual power plant according to the charging and discharging power reference value. By constructing the distributed energy cluster and multi-element energy storage collaborative optimization control method, efficient and stable operation of the virtual power plant is achieved, and the energy utilization rate and the power grid supporting capacity are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual power plant control, and provides a virtual power plant control method, system and device based on the coordination of energy clusters and energy storage. Background Art

[0002] Under the existing separate control mode of distributed energy, solar photovoltaic power stations, wind farms, etc. act independently and lack an effective coordination mechanism. When sunlight is suddenly blocked by clouds, the power generation power of the photovoltaic power station drops sharply. At this time, if the wind farm cannot supplement the power gap in time due to too low wind speed, and there is no other energy or energy storage system in the surrounding area for regulation, it will lead to unstable power supply in local areas. When the light is sufficient and the wind is strong, each energy overgenerates electricity separately, but they cannot cooperate with each other to reasonably store or transport the excess electric energy, resulting in a large amount of energy being wasted in the end.

[0003] A single energy storage control method greatly limits the performance of the energy storage system. Taking lithium batteries as an example, in some industrial production scenarios with sudden high-power electricity consumption, the energy storage system needs to quickly release a large amount of electric energy. Due to its own characteristics, lithium batteries are difficult to maintain good performance during high-power fast charge and discharge processes, and problems such as unstable voltage and shortened lifespan may occur. Although supercapacitors can quickly respond to high-power demands, their energy storage capacity is limited and they cannot meet long-term and large-scale energy storage needs. This makes the energy storage system appear powerless in the face of complex and diverse energy fluctuations.

[0004] Traditional centralized control architectures have obvious drawbacks in the application of virtual power plants. Due to the wide distribution of distributed energy and energy storage devices, a large amount of data needs to be transmitted to the central controller for processing. Information will inevitably be delayed during the transmission process. Coupled with the fact that the central controller also needs time to process data, when the grid load suddenly increases, or the power generation power of distributed energy changes significantly due to weather and other factors, it is difficult for the central controller to issue accurate and effective control instructions in a timely manner, and it is unable to quickly adjust the energy distribution and the working state of the energy storage system, seriously affecting the response ability of the virtual power plant to grid dynamic changes. Summary of the Invention

[0005] The present invention aims to at least solve one of the technical problems existing in the related technologies. Therefore, the present invention provides a virtual power plant control method, system and device based on the coordination of energy clusters and energy storage, effectively solving the problems of distributed energy acting independently and insufficient coordination, avoiding energy waste, ensuring power supply stability and reliability, and improving energy utilization efficiency.

[0006] The present invention provides a virtual power plant control method based on the coordination of energy clusters and energy storage, including: S1: Collect real-time data of energy devices, and preprocess the real-time data to obtain preprocessed real-time data; S2: Use the LSTM attention graph neural network to perform predictions on the preprocessed real-time data to obtain prediction results; S3: Based on the prediction results, configure and co-optimize the multi-energy storage system to obtain the reference charge and discharge power values; S4: Adjust the generator sets in the virtual power plant according to the reference charge and discharge power values.

[0007] According to a virtual power plant control method based on energy cluster and energy storage collaboration provided by the present invention, the real-time data includes power generation power, equipment status, and environmental parameters.

[0008] According to a virtual power plant control method based on energy cluster and energy storage collaboration provided by the present invention, the preprocessing method includes: S11: Sort the collected real-time data in chronological order to obtain the original real-time data; S12: Filter out the original real-time data with a length less than the length threshold to obtain the filtered original real-time data; S13: Fill in the missing values in the filtered original real-time data to obtain the preprocessed real-time data.

[0009] According to a virtual power plant control method based on energy cluster and energy storage collaboration provided by the present invention, step S2 includes: S21: Optimize the input preprocessed real-time data according to the dynamic gated convolution to obtain the optimized real-time data; S22: Use the LSTM algorithm to predict the preprocessed real-time data to obtain the hidden state; S23: Use the graph neural network and attention mechanism to dynamically allocate and optimize the hidden state to obtain the prediction result.

[0010] According to a virtual power plant control method based on energy cluster and energy storage collaboration provided by the present invention, step S21 includes: S211: Optimize the prediction input using the dynamic gated convolution: where, is the final convolution kernel, is the first activation function, is the multi-layer perceptron function, is the preprocessed real-time data at the past time moments, is the current moment, is the backtracking time step, is the window start time, is the initial convolution kernel; S212: Calculate the learnable dilation rate : Among them, is the rounding function, is the inflation rate learning parameter, is the predicted scaling factor function, represents the preprocessed real-time data at the current moment, is the maximum inflation rate; S213: Calculate the channel gating: Among them, is the channel gating of the th energy device, is the ordinal number of the energy device, , is the total number of energy devices, is the normalized weight parameter, is the global average pooling function; S214: Fuse the LSTM input: Among them, is the LSTM additional value, is the one-dimensional convolution function, is the th energy device's , is the th energy device's , is the optimized real-time data.

[0011] According to a virtual power plant control method based on energy cluster and energy storage cooperation provided by the present invention, step S23 includes: S231: Optimize the hidden state using a graph neural network: Among them, is the updated hidden state of the power station , is the second activation function, is the neighbor power station of the power station , is the set of neighbor power stations of the power station , is the degree of the power station , is the degree of the power station ; is the learnable matrix; is the neighbor power station The original hidden state at time t; S232: Calculate the importance score for each time step: where, is the importance score at the current time, is the initial attention weight matrix, represents matrix transpose, is the hyperbolic tangent activation function, is the trainable weight, is the bias, is the updated hidden state; S233: Calculate the context vector : where, is the context weight, is the importance score of the neighboring power plants; S234: Combine the context vector with the updated hidden state to obtain the prediction result: where, is the prediction result, is the output trainable weight, is the output bias.

[0012] According to a virtual power plant control method based on energy cluster and energy storage cooperation provided by the present invention, step S3 includes: S31: Establish dynamic equations of different SOCs according to different energy storage technologies; S32: Perform MPC-RL modeling on the power grid system to obtain the MPC-RL algorithm; S33: Input the dynamic equation of the SOC and the prediction result into the MPC-RL algorithm to solve for the reference charge and discharge power value.

[0013] The present invention also provides a virtual power plant control system based on energy cluster and energy storage cooperation, including: Data preprocessing module: Collect real-time data of energy devices and preprocess the real-time data to obtain preprocessed real-time data; Data prediction module: Use the LSTM attention graph neural network to perform prediction based on the preprocessed real-time data to obtain a prediction result; Data cooperation optimization module: Perform multi-energy storage system configuration and cooperation optimization based on the prediction result to obtain the reference charge and discharge power value; Generator set regulation module: regulate the generator sets in the virtual power plant according to the charge-discharge power reference value.

[0014] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the virtual power plant control method based on energy cluster and energy storage collaboration as described in any one of the above are implemented.

[0015] One or more of the above technical solutions in the embodiments of the present invention have at least one of the following technical effects: A virtual power plant control method, system, and device based on energy cluster and energy storage collaboration provided by the present invention, through the application of Internet of Things technology and LSTM model and adding an attention mechanism to the model, collect distributed energy data in real time, accurately predict it, construct a global optimization model, and realize reasonable energy distribution. It effectively solves the problems of distributed energy acting independently and insufficient collaboration, avoids energy waste, ensures power supply stability and reliability, and improves energy utilization efficiency.

[0016] Establish accurate mathematical models of different energy storage technologies, adopt a hierarchical control strategy, globally optimize with the upper-layer MPC-RL algorithm, and locally precisely control each energy storage device in the lower layer. Considering grid power fluctuations, energy storage charge-discharge smoothness, and battery health status, it improves the adaptability of the energy storage system under different time scales and power demands, and fully exerts the characteristics of different energy storage technologies.

[0017] Optimize and control with the goal of minimizing the overall operating cost of the virtual power plant and maximizing energy utilization efficiency. By reasonably arranging distributed energy generation and energy storage system charge-discharge, it reduces energy loss and operating cost, and improves the economic benefits of the virtual power plant.

[0018] The additional aspects and advantages of the present invention will be partially given in the following description, partially become obvious from the following description, or be understood through the practice of the present invention. Description of the Drawings

[0019] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0020] Figure 1 It is a schematic flowchart of the virtual power plant control method based on energy cluster and energy storage collaboration provided by the present invention.

[0021] Figure 2It is a structural block diagram of a control device for a virtual power plant based on the collaboration of an energy cluster and energy storage provided by the present invention.

[0022] Figure 3 It is a schematic structural diagram of an electronic device provided by the present invention.

[0023] Reference numerals: 101, data preprocessing module; 102, data prediction module; 103, data collaborative optimization module; 104, generator unit regulation module; 810, processor; 820, communication interface; 830, memory; 840, communication bus. Detailed implementation manners

[0024] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments in the present invention belong to the scope of protection of the present invention. The following embodiments are used to illustrate the present invention, but cannot be used to limit the scope of the present invention.

[0025] In the description of this specification, the descriptions referring to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the embodiments of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, without conflict, those skilled in the art can combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0026] The following is combined with Figures 1 to 3 to describe the present invention.

[0027] Embodiment As Figure 1 shown, Figure 1 It is a schematic flow diagram of a control method for a virtual power plant based on the collaboration of an energy cluster and energy storage, including the following steps: S1: Collect real-time data of energy devices, and preprocess the real-time data to obtain preprocessed real-time data; S2: Use the LSTM attention graph neural network to perform prediction based on the preprocessed real-time data to obtain a prediction result; S3: Perform configuration and collaborative optimization of a multi-energy storage system based on the prediction result to obtain a reference value for charging and discharging power; S4: Adjust the generating units in the virtual power plant according to the charging and discharging power reference value.

[0028] Specifically, the real-time data includes generating power, equipment status, and environmental parameters. With the Internet of Things technology, a large number of sensors are deployed on distributed energy equipment (such as photovoltaic panels, wind turbines, etc.) to collect real-time data such as generating power, equipment status, and environmental parameters (light intensity, wind speed, temperature, etc.). The data is transmitted to the central control platform through a wireless communication network (such as ZigBee, LoRa, etc.).

[0029] Specifically, the preprocessing method includes: S11: Sort the collected real-time data in chronological order to obtain the original real-time data; S12: Filter out the original real-time data with a length less than the length threshold to obtain the filtered original real-time data; S13: Fill in the missing values in the filtered original real-time data to obtain the preprocessed real-time data.

[0030] Specifically, step S2 includes: S21: Optimize the input preprocessed real-time data according to the dynamic gated convolution to obtain the optimized real-time data.

[0031] Step S21 includes: S211: Optimize the prediction input using the dynamic gated convolution: Among them, is the final convolution kernel, is the first activation function, is the multi-layer perceptron function, is the past preprocessed real-time data at current time, backtracking time step, window start time; S212: Calculate the learnable dilation rate : Among them, is the rounding function, is the dilation rate learning parameter, is the prediction scaling factor function, represents the preprocessed real-time data at the current time, is the maximum dilation rate; S213: Calculate the channel gating: Among them, For the channel gating of the th energy device, is the ordinal number of the energy device, , is the total number of energy devices, is the normalized weight parameter, is the global average pooling function; S214: Fuse the LSTM input: Among them, is the LSTM additional value, is the one-dimensional convolution function, For the th energy device's , For the th energy device's , is the optimized real-time data.

[0032] S22: Use the LSTM algorithm to predict the preprocessed real-time data to obtain the hidden state. Step S22 includes: S221: Determine which information should be added to the current memory: In the formula, is the input gate state, which determines which information should be added to the current memory. The sigmoid activation function is used to compress the input between 0 and 1 to control the information flow, is the concatenation of the previous hidden state and the current input, is the input gate bias term, is the input gate weight matrix.

[0033] S222: Determine which states should be discarded from the old memory: In the formula is the forget gate state, is the forget gate weight matrix, is the forget gate bias term.

[0034] S223: Determine which information should be output to the next time step: In the formula is the output gate state, is the output gate weight matrix, is the output gate bias term.

[0035] S224: Calculate the hidden state: is a convolution operation, is the hidden state at the previous moment, Candidate memory interest matrix, is the candidate memory bias term, is element-wise multiplication.

[0036] S23: Dynamically allocate and optimize the hidden state using a graph neural network and an attention mechanism to obtain a prediction result. Step S23 includes: S231: Optimize the hidden state using a graph neural network: where, is the updated hidden state of the power station , is the second activation function, is the power station 's neighboring power station, is the power station 's set of neighboring power stations, is the power station 's degree, is the power station 's degree; is a learnable matrix; is the neighboring power station 's original hidden state at time t; S232: Calculate the importance score for each time step: where, is the importance score at the current moment, is the initial attention weight matrix, represents matrix transpose, is the hyperbolic tangent activation function, is the trainable weight, is the bias term, is the updated hidden state; S233: Calculate the context vector : where, is the context weight, is the importance score of the neighboring power station; S234: Combine the context vector with the updated hidden state to obtain the prediction result: where is the prediction result, is the output trainable weight, is the output bias.

[0037] Bring an example into the model for calculation: Task: Predict the photovoltaic power generation at 12:00 noon tomorrow.

[0038] Input data: Illumination intensity in the past 3 hours (one data point per hour): (Normalized values, 0 means no illumination, 1 means the strongest illumination), where is the schematic input data group, is the first schematic data, is the second schematic data, is the third schematic data.

[0039] LSTM hidden state (assuming it has been calculated through forward propagation, representing the change trend of illumination intensity and the device state) is the first schematic hidden state, is the second schematic hidden state, is the third schematic hidden state.

[0040] Step 1: Calculate the attention scores Concatenate the input and the hidden state: Calculate the scores: where is the first schematic importance score, is the second schematic importance score, is the third schematic importance score.

[0041] Step 2: Normalize the attention weights ; ; where is the first schematic context weight, the second schematic context weight, is the third schematic context weight.

[0042] Step 3: Weightedly aggregate all hidden states to generate a context vector ; First dimension: ; Second dimension: ; Finally obtain: ; Step 4: Final prediction output ; Assume: , Traditional LSTM calculation process: .

[0043] After adding the attention mechanism, the core advantage of the LSTM model is that it can dynamically focus on key historical information, significantly improving the accuracy and robustness of prediction. Traditional LSTM treats all historical data equally, while the attention mechanism calculates weights automatically, preferentially extracting the features most relevant to the current prediction and ignoring irrelevant noise.

[0044] By applying Internet of Things technology and the LSTM model, adding the attention mechanism and graph neural network algorithm to the model, real-time collection and accurate prediction of distributed energy data are carried out, a global optimization model is constructed, and reasonable energy distribution is achieved. It effectively solves the problems of decentralized operation and insufficient coordination of distributed energy, avoids energy waste, ensures power supply stability and reliability, and improves energy utilization efficiency.

[0045] Specifically, step S3 includes: S31: Establish dynamic equations of different SOCs according to different energy storage technologies; Specifically, taking lithium batteries as an example, the dynamic equation of its state of charge SOC is: Among them, is the derivative of the state of charge SOC with respect to time, reflecting the rate of change of the state of charge over time. I is the battery charge and discharge current, and Q is the battery rated capacity. For supercapacitors, the relationship between its voltage V and the battery rated capacity Q is , where C is the capacitance value.

[0046] S32: Conduct MPC-RL modeling on the power grid system to obtain the MPC-RL algorithm, including the following steps: S321: Upper-layer global optimization: An optimization model is established with the objective function of minimizing the overall operating cost of the virtual power plant and maximizing the energy utilization rate, considering constraints such as the charge-discharge power limit and state of charge limit of the energy storage system. The objective function can be expressed as: where, is the distributed energy generation cost at time , is the operating cost of the energy storage system at time , denotes taking the minimum value, is the total duration.

[0047] The constraints include: (Power balance constraint); (State of charge constraint); (Energy storage discharge power constraint); (Energy storage charge power constraint); In the formula, is the total power generation at time , is the discharge power of the energy storage system at time , is the charge power of the energy storage system at time , is the load power at time , is the minimum state of charge value, is the state of charge value at time is the maximum state of charge value, is the minimum energy storage system discharge power value, is the maximum energy storage system discharge power value, is the minimum energy storage system charge power value, is the maximum energy storage system charge power value.

[0048] The model predictive control (MPC) method is used to solve this optimization problem, and the reference values of the discharge power and charge power of the energy storage system at a future time are obtained.

[0049] S323: Each energy storage device performs local control based on the reference values issued by the upper layer and its own state. For example, the lithium battery energy storage system realizes charge control by adjusting the current of the charging circuit according to the reference charge power; the supercapacitor quickly releases energy according to the reference discharge power.

[0050] S33: Input the dynamic equation of the SOC and the prediction result into the MPC-RL algorithm to solve for the reference charge and discharge power value, including the following steps: S331: Define variables: Set the control variable : Set the state variable : RL_state(T) represents the auxiliary state provided by RL: Prediction error : In the formula, is the actual power generation value at time t, is the predicted power generation value at time t, represents the vector norm.

[0051] Grid fluctuation: ; Historical control command ; RL dynamic weight : ; : The dynamic weight of the distributed energy generation cost (0 - 1), adjusted by RL according to the electricity price and prediction error; : The dynamic weight of the energy storage operation cost (0 - 1), adjusted by RL according to the SOC and grid stability.

[0052] S332: Model according to the optimization problem: The corrected objective function The expression is: Among them, represents the control variable corresponding to the minimum value, is the penalty term generated by RL.

[0053] Among them, is the extreme SOC coefficient, is the frequency fluctuation coefficient.

[0054] S333: Collaborative solution process: The RL module calculates in real time: Input: Prediction result ; Output: Dynamic weight λ; MPC Solving: Input: New optimization function, constraint conditions, and dynamic weight λ; Output: Optimal control sequence; Finally, perform closed-loop solving, repeat the process, and solve the charging and discharging strategy that takes into account economy and stability under the guidance of RL.

[0055] Establish an accurate mathematical model for different energy storage technologies, adopt a hierarchical control strategy, globally optimize the upper-layer MPC-RL algorithm, and locally and precisely control each energy storage device in the lower layer. Taking into account the power fluctuations of the power grid, the smoothness of energy storage charging and discharging, and the battery health status, the adaptability of the energy storage system under different time scales and power demands is improved, and the characteristics of different energy storage technologies are fully utilized.

[0056] Optimal control is carried out with the goal of minimizing the overall operating cost of the virtual power plant and maximizing the energy utilization rate. By reasonably arranging the distributed energy generation and the charging and discharging of the energy storage system, the energy loss and operating cost are reduced, and the economic benefits of the virtual power plant are improved.

[0057] Table 1 Comparison Table between the Present Invention and Traditional Methods As shown in Table 1, Table 1 is a comparison table between the present invention and traditional methods, and it can be seen that the present invention has great advantages over traditional methods.

[0058] As Figure 2 shown, the present invention also provides a virtual power plant control system based on the coordination of energy clusters and energy storage, including: Data preprocessing module 101: Collect real-time data of energy devices and preprocess the real-time data to obtain preprocessed real-time data; Data prediction module 102: Use the LSTM attention graph neural network to perform prediction based on the preprocessed real-time data to obtain a prediction result; Data collaborative optimization module 103: Perform multi-energy storage system configuration and collaborative optimization based on the prediction result to obtain a reference value for charging and discharging power; Generator set adjustment module 104: Adjust the generator sets in the virtual power plant according to the reference value of the charging and discharging power.

[0059] Figure 3 Illustrates a schematic diagram of the physical structure of an electronic device. As Figure 3 shown, the electronic device may include: a processor 810, a communication interface 820, a memory 830, and a communication bus 840. Among them, the processor 810, the communication interface 820, and the memory 830 complete mutual communication through the communication bus 840. The processor 810 can call the logical instructions in the memory 830 to execute a virtual power plant control method based on the coordination of energy clusters and energy storage. The method includes: S1: Collect the real-time data of the energy equipment, and preprocess the real-time data to obtain preprocessed real-time data; S2: Use the LSTM attention graph neural network to make predictions based on the preprocessed real-time data to obtain prediction results; S3: Configure and co-optimize the multi-energy storage system according to the prediction results to obtain the reference charge and discharge power values; S4: Adjust the generator sets in the virtual power plant according to the reference charge and discharge power values.

[0060] In addition, when the logic instructions in the above-mentioned memory 830 are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0061] On the other hand, the present invention also provides a computer program product, where the computer program product includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the control method for a virtual power plant based on energy cluster and energy storage collaboration provided by the above-mentioned various methods.

[0062] On yet another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it is configured to execute the control method for a virtual power plant based on energy cluster and energy storage collaboration provided by the above-mentioned various methods.

[0063] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0064] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, 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 ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0065] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0066] It should be noted that the embodiments of the present disclosure can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic: the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art can understand that the above devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code is provided on a programmable memory or a data carrier such as an optical or electronic signal carrier.

[0067] In addition, although the operations of the method of the present disclosure are described in a specific order in the drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. On the contrary, the steps depicted in the flowchart can be changed in the order of execution. Additionally or alternatively, some steps can be omitted, multiple steps can be combined into one step for execution, and / or one step can be decomposed into multiple steps for execution. It should also be noted that the features and functions of two or more devices according to the present disclosure can be embodied in one device. Conversely, the features and functions of one device described above can be further divided and embodied by multiple devices.

[0068] Although the present disclosure has been described with reference to several specific embodiments, it should be understood that the present disclosure is not limited to the specific embodiments disclosed. The present disclosure aims to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

Claims

1. A control method for a virtual power plant based on the coordination of an energy cluster and energy storage, characterized in that, Including: S1: Collect the real-time data of energy equipment, and preprocess the real-time data to obtain preprocessed real-time data. The preprocessing method includes: S11: Sort the collected real-time data in chronological order to obtain the original real-time data; S12: Filter out the original real-time data with a length less than the length threshold to obtain the filtered original real-time data; S13: Fill in the missing values of the filtered original real-time data to obtain the preprocessed real-time data; S2: Use the LSTM attention graph neural network to perform prediction based on the preprocessed real-time data to obtain a prediction result. Step S2 includes: S21: Optimize the input preprocessed real-time data according to the dynamic gated convolution to obtain the optimized real-time data; S22: Use the LSTM algorithm to predict the preprocessed real-time data to obtain the hidden state; S23: Use the graph neural network and the attention mechanism to dynamically allocate and optimize the hidden state to obtain the prediction result; S3: Perform multi-energy storage system configuration and collaborative optimization based on the prediction result to obtain the charge and discharge power reference value; S4: Adjust the generator sets in the virtual power plant according to the charge and discharge power reference value.

2. The control method of the virtual power plant based on the cooperation of the energy cluster and energy storage according to claim 1, characterized in that The real-time data includes power generation power, equipment status, and environmental parameters.

3. The control method of the virtual power plant based on the cooperation of the energy cluster and energy storage according to claim 1, wherein Step S21 includes: S211: Optimize the prediction input using the dynamic gated convolution: Among them, is the final convolution kernel, is the first activation function, is the multi-layer perceptron function, is the preprocessed real-time data for the past time instants, is the current time instant, is the backtracking time step, is the window start time, is the initial convolution kernel; S212: Calculate the learnable dilation rate : Among them, is the rounding function, is the inflation rate learning parameter, is the predicted scaling factor function, represents the preprocessed real-time data at the current moment, is the maximum inflation rate; S213: Calculate the channel gate control: Among them, is the channel gating of the th energy device, is the ordinal number of the energy device, , is the total number of energy devices, is the normalized weight parameter, is the global average pooling function; S214: Integrate the LSTM input: Among them, is the added value of LSTM, is a one-dimensional convolution function, is the of the th energy device, is the of the th energy device, is the optimized real-time data.

4. The control method of the virtual power plant based on the cooperation of the energy cluster and energy storage according to claim 3, characterized in that, Step S23 includes: S231: Optimize the hidden state using the graph neural network: Among them, is the updated power station in the hidden state, is the second activation function, is the power station of the neighboring power station, is the power station a set of neighboring power stations, is the power station the degree of, is the power station the degree of; is a learnable matrix; is the neighboring power station in the original hidden state at time t; S232: Calculate the importance score for each time step: Among them, is the importance score at the current moment, is the initial attention weight matrix, represents matrix transpose, is the hyperbolic tangent activation function, is the trainable weight, is the bias, is the updated hidden state; S233: Calculate the context vector : Among them, is the context weight, is the importance score of the neighboring power station; S234: Combine the context vector with the updated hidden state to obtain the prediction result: Among them, is the prediction result, is the output trainable weight, is the output bias.

5. The control method of the virtual power plant based on the coordination of the energy cluster and energy storage according to claim 1, wherein, Step S3 includes: S31: Establish dynamic equations for different SOCs according to different energy storage technologies; S32: Perform MPC-RL modeling on the power grid system to obtain the MPC-RL algorithm; S33: Input the dynamic equation of the SOC and the prediction result into the MPC-RL algorithm to solve and obtain the charge and discharge power reference value.

6. A control system for an energy cluster and energy storage collaborative virtual power plant, which is used to execute the control method for an energy cluster and energy storage collaborative virtual power plant according to any one of claims 1 to 5, characterized in that, Including: Data preprocessing module: Collect the real-time data of energy equipment, and preprocess the real-time data to obtain preprocessed real-time data; Data prediction module: Use the LSTM attention graph neural network to perform prediction based on the preprocessed real-time data to obtain a prediction result; Data collaborative optimization module: Perform multi-energy storage system configuration and collaborative optimization based on the prediction result to obtain the charge and discharge power reference value; Generator set adjustment module: Adjust the generator sets in the virtual power plant according to the charge and discharge power reference value.

7. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the virtual power plant control method based on the cooperation of the energy cluster and energy storage as described in any one of claims 1 to 5.

Images