A method, device and medium for evaluating adjustable capacity of a multi-energy coupled virtual power plant

Abstract

The present application discloses a method, device and medium for evaluating the adjustable capacity of a multi-energy coupled virtual power plant, which is used to solve the problem that the prior art ignores the coupling relationship and mutual influence between multiple energy sources when constructing an adjustable capacity evaluation strategy for a multi-energy coupled virtual power plant, and is insufficient in predicting the output of uncertain resources. The method includes: constructing a multi-energy coupled resource model within a multi-energy coupled virtual power plant to determine the resource type within the multi-energy coupled virtual power plant; the resource type includes uncertain adjustable resources and deterministic adjustable resources; based on a deep learning network, constructing an output prediction model for uncertain adjustable resources to determine the output prediction results corresponding to uncertain adjustable resources; constructing an evaluation strategy for the adjustable capacity of resources corresponding to the multi-energy coupled virtual power plant, and combining the output prediction results corresponding to the uncertain adjustable resources and the deterministic adjustable resources to determine the adjustable capacity corresponding to the multi-energy coupled virtual power plant.

Description

Technical Field

[0001] The present application relates to the technical field of adjustable capacity assessment of virtual power plants, and in particular to an adjustable capacity assessment method, device and medium for a multi-energy coupled virtual power plant. Background Art

[0002] At present, with the transformation of the global energy structure and the rapid development of smart grid technology, multi-energy coupled virtual power plants (VPPs) as an innovative energy management model are gradually becoming an important means to solve the efficient use of energy, flexible grid dispatch and improve system stability. Traditional power systems mainly focus on the single dimension of electric energy, while multi-energy coupled virtual power plants break this limitation and expand the perspective to the integrated energy system, realizing the deep integration and coordinated management of multiple energy forms such as electricity, heat energy, and cold energy.

[0003] In modern power systems, the disorderly power consumption of a large number of demand-side response resources, such as adjustable loads and energy storage systems, not only directly threatens the safe and stable operation of the power grid, but may also cause unnecessary economic losses to users due to inappropriate scheduling strategies. However, these resources actually contain huge regulatory flexibility, and through effective management and scheduling, the flexibility and reliability of the power system can be significantly improved. The concept of virtual power plant came into being. It integrates these scattered demand-side response resources through advanced information and communication technology and intelligent management systems to form a virtual power generation unit that can be uniformly scheduled and optimized, realizing flexible conversion of the roles of power generation and power consumption.

[0004] The advantage of a multi-energy coupled virtual power plant is that it can comprehensively consider the complementarity and synergy between various energy forms, and achieve efficient utilization and flexible configuration of energy by optimizing the dispatching strategy. For example, during peak hours of electricity demand, a virtual power plant can dispatch energy storage systems to discharge or use distributed renewable energy to generate electricity. At the same time, through the conversion and storage of heat or cold energy, it can meet the diversified energy needs of users, thereby reducing the power supply pressure of the power grid. In addition, a multi-energy coupled virtual power plant can also respond to market price signals and grid dispatching instructions in real time through an intelligent management system, achieving a win-win situation in economic and environmental benefits.

[0005] However, existing technologies still have many deficiencies in constructing their adjustable capacity evaluation strategies and clarifying the operating boundary conditions. Existing evaluation methods often focus on the analysis of the adjustable capacity of a single energy form, ignoring the coupling relationship and mutual influence between multiple energy forms, resulting in incomplete and inaccurate evaluation results. At the same time, when predicting the output of uncertain resources (such as renewable energy generation), traditional methods often find it difficult to accurately capture their volatility and randomness, and thus cannot provide a reliable basis for the optimal scheduling of virtual power plants. Summary of the invention

[0006] The embodiments of the present application provide a method, device and medium for evaluating the adjustable capacity of a multi-energy coupled virtual power plant, so as to solve the technical problems that the prior art ignores the coupling relationship and mutual influence among multiple energy sources when constructing the adjustable capacity evaluation strategy of the multi-energy coupled virtual power plant, and is insufficient in predicting the output of uncertain resources.

[0007] On the one hand, an embodiment of the present application provides a method for evaluating the adjustable capacity of a multi-energy coupled virtual power plant, including:

[0008] Constructing a multi-energy coupling resource model in a multi-energy coupling virtual power plant to determine the resource types in the multi-energy coupling virtual power plant; the resource types include uncertain adjustable resources and deterministic adjustable resources;

[0009] Based on the deep learning network, an output prediction model of an uncertain adjustable resource is constructed to determine the output prediction result corresponding to the uncertain adjustable resource;

[0010] A resource adjustable capacity evaluation strategy corresponding to the multi-energy coupling virtual power plant is constructed, and the adjustable capacity corresponding to the multi-energy coupling virtual power plant is determined by combining the output prediction results corresponding to the uncertain adjustable resources and the deterministic adjustable resources.

[0011] In one implementation of the present application, the uncertain adjustable resources include distributed power sources and uncertain loads;

[0012] The method further comprises:

[0013] The uncertain load is divided into non-adjustable load, curtailable load, transferable load and shiftable load, and the power consumption corresponding to the uncertain load is obtained to generate a corresponding power consumption curve;

[0014] The proportional relationship among the reducible load, the transferable load and the shiftable load is determined, and based on a preset maximum power consumption threshold and a minimum power consumption threshold, an adjustable space corresponding to the uncertain load is determined.

[0015] In one implementation of the present application, the uncertainty adjustable resource also includes an electric vehicle charging station;

[0016] The method further comprises:

[0017] Acquiring historical data of the electric vehicle charging station, and determining the dispatching potential of the electric vehicle charging station based on the historical data;

[0018] The historical data is expressed by the following formula:

[0019]

[0020] in, represents the historical data of the electric vehicle charging station, Indicates electric vehicle Arrival time, Indicates the electric vehicle The departure time, Indicates the electric vehicle The battery level reaches Indicates the electric vehicle The battery power is left, Indicates the electric vehicle The upper limit of battery power, Indicates the electric vehicle The battery power limit is Indicates the electric vehicle The upper limit of charging power, Indicates the electric vehicle The lower limit of charging power.

[0021] In one implementation of the present application, the deterministic adjustable resources include energy storage power stations and gas turbines;

[0022] The method further comprises:

[0023] Monitor the energy storage power station and obtain the charging and discharging status corresponding to the energy storage power station to determine the adjustable range corresponding to the energy storage power station;

[0024] The gas turbine is monitored and the corresponding operating status of the gas turbine is obtained to determine the corresponding adjustable range of the gas turbine.

[0025] In one implementation of the present application, the uncertainty adjustable resource output prediction model is constructed based on the deep learning network to determine the output prediction result corresponding to the uncertainty adjustable resource, specifically including:

[0026] Based on the deep learning network, an uncertain adjustable resource output prediction model is constructed, and sample data is obtained to perform standardization processing on the sample data;

[0027] Inputting the processed sample data into the uncertainty adjustable resource output prediction model to encode the sample data through an activation function to obtain corresponding hidden layer variables;

[0028] The hidden layer variables are as follows:

[0029]

[0030]

[0031]

[0032] in, Represents the input matrix of processed sample data, represents the parameter matrix after being encoded by the activation function, Represents the total number of inputs, represents the linear relationship coefficient, Represents the error bias between the parameter matrix and the input matrix.

[0033] In one implementation of the present application, the method further includes:

[0034] The output after multiple hidden layers is expressed by the following formula: And the decoded function The output layer The mathematical relationship between:

[0035]

[0036]

[0037]

[0038] in, Represents the output matrix after multiple hidden layers, Indicates that the decoded function The output layer parameter matrix, represents the linear relationship coefficient between the two layers, Represents the error bias between the output layer parameter matrix and the output quantity matrix.

[0039] In one implementation of the present application, determining the proportional relationship between the reducible load, the transferable load, and the shiftable load, and determining the adjustable space corresponding to the uncertain load based on a preset maximum power consumption threshold and a minimum power consumption threshold, specifically includes:

[0040] The corresponding down-regulation range of distributed power generation is calculated by the following formula:

[0041]

[0042] in, represents the downward adjustment range of the distributed power source, Indicates that the distributed power source is The predicted output at the time;

[0043] The output operation downward adjustment boundary and output operation upward adjustment boundary of the electric vehicle charging station are calculated by the following formula:

[0044]

[0045]

[0046] in, Indicates that the electric vehicle charging station The lower limit of the moment, Indicates that the electric vehicle charging station The upward adjustment limit of the moment, represents the forecast data of electric vehicle charging stations, represents the lower bound of the operating capacity of the electric vehicle charging station, represents the upper limit of the operating capacity of the electric vehicle charging station;

[0047] The adjustable capacity of uncertain load is calculated by the following formula:

[0048]

[0049]

[0050] in, Indicates that the uncertainty load is The load at the moment reduces the power. Indicates that the uncertainty load is The load at the moment increases the power. Indicated in The power reduction that can be achieved by reducing the load at any time, Indicated in The variable power of the load that can be transferred at any time, Indicated in The load can be shifted at any time. Indicated in The maximum output of the total load at any time, Indicated in The current output of the load at that moment.

[0051] In one implementation of the present application, the resource adjustable capacity evaluation strategy corresponding to the multi-energy coupling virtual power plant is constructed, and the adjustable capacity corresponding to the multi-energy coupling virtual power plant is determined by combining the output prediction results corresponding to the uncertain adjustable resources and the deterministic adjustable resources, specifically including:

[0052] The adjustable power of the energy storage device is calculated using the following formula:

[0053]

[0054]

[0055] in, and Indicated in The adjustable capacity of the energy storage device at the time, Indicated in The SOC state of the energy storage device at the moment, represents the lower bound of the capacity of the energy storage device, represents the upper limit of the capacity of the energy storage device;

[0056] The adjustable power of the gas turbine is calculated by the following formula:

[0057]

[0058]

[0059] in, and Indicated in The adjustable power of the gas turbine at the moment, Indicated in The power generation capacity of the gas turbine at the time, represents the lower limit of the gas turbine output, represents the upper limit of the gas turbine output.

[0060] On the other hand, an embodiment of the present application further provides an adjustable capacity evaluation device for a multi-energy coupled virtual power plant, the device comprising:

[0061] at least one processor;

[0062] and, a memory communicatively coupled to the at least one processor;

[0063] Wherein, the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the adjustable capacity assessment method of a multi-energy coupled virtual power plant as described above.

[0064] On the other hand, an embodiment of the present application further provides a non-volatile computer storage medium storing computer executable instructions, which, when executed, implements a method for evaluating the adjustable capacity of a multi-energy coupled virtual power plant as described above.

[0065] The present application provides a method, device and medium for evaluating the adjustable capacity of a multi-energy coupled virtual power plant, which at least have the following beneficial effects:

[0066] By clearly distinguishing between uncertain adjustable resources and deterministic adjustable resources, virtual power plants can manage and dispatch various types of resources more accurately, thereby improving the efficiency and flexibility of resource utilization; by constructing a multi-energy coupling resource model, virtual power plants can better adapt to the characteristics and changes of different resource types, and improve their adaptability to complex and changeable energy environments; through deep learning technology, it is possible to deeply mine the implicit information in the data, thereby more accurately predicting the output of uncertain adjustable resources, providing more reliable data support for the dispatch and decision-making of virtual power plants; through accurate output prediction results, virtual power plants can be more scientific and reasonable when formulating dispatch plans and decisions, thereby improving the economy and stability of their operation; by evaluating the adjustable capacity of various resources, virtual power plants can formulate dispatch plans more flexibly to cope with energy demands and changes in different scenarios; by combining output prediction results and resource adjustable capacity assessment, virtual power plants can allocate and utilize various resources more scientifically, thereby improving the economy and environmental protection of their operation, and achieving a win-win situation of social and economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0068] Figure 1 A flow chart of a method for evaluating the adjustable capacity of a multi-energy coupled virtual power plant provided in an embodiment of the present application;

[0069] Figure 2 A schematic diagram of the deep neural network structure provided in the embodiment of the present application;

[0070] Figure 3 A schematic diagram of an offline training framework of a DNN model based on an improved Adam optimizer provided in an embodiment of the present application;

[0071] Figure 4 A schematic diagram of the internal structure of an adjustable capacity assessment device for a multi-energy coupled virtual power plant provided in an embodiment of the present application. DETAILED DESCRIPTION

[0072] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0073] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.

[0074] Figure 1 A flow chart of a method for evaluating the adjustable capacity of a multi-energy coupled virtual power plant provided in an embodiment of the present application.

[0075] The analysis method involved in the embodiments of the present application can be implemented by a terminal device or a server, and the present application does not impose any special restrictions on this. For the convenience of understanding and description, the following embodiments are described in detail by taking a server as an example.

[0076] It should be noted that the server may be a single device or a system consisting of multiple devices, that is, a distributed server, and this application does not make any specific limitation on this.

[0077] like Figure 1 As shown, an embodiment of the present application provides a method for evaluating the adjustable capacity of a multi-energy coupled virtual power plant, including:

[0078] 101. Construct a multi-energy coupling resource model within a multi-energy coupling virtual power plant to determine the resource types within the multi-energy coupling virtual power plant.

[0079] It should be noted that the resource types in the embodiments of the present application include uncertain adjustable resources and deterministic adjustable resources.

[0080] Specifically, uncertain adjustable resources include distributed power sources and uncertain loads. Distributed power sources include distributed wind power and photovoltaic power. Since the output of distributed power sources is highly intermittent and random, it is not only affected by climatic conditions such as wind, light, season, and operating conditions such as geography and time, but also by adjacent power sources and power loads. Therefore, accurate modeling and characteristic analysis of distributed power sources is a complex mathematical problem involving multiple couplings and multiple influencing factors.

[0081] The server divides the uncertain load into non-adjustable load, curtailable load, transferable load and shiftable load, and obtains the power consumption corresponding to the uncertain load to generate a corresponding power consumption curve; determines the proportional relationship between the curtailable load, the transferable load and the shiftable load, and determines the adjustable space corresponding to the uncertain load based on the preset maximum power consumption threshold and minimum power consumption threshold.

[0082] It should be noted that curtailable loads, transferable loads and shiftable loads are adjustable resources. Curtailable loads are loads that can be temporarily reduced or interrupted within the response time window. Such loads usually come from non-critical electrical equipment that does not affect the user's core business or quality of life. Taking air conditioning load as an example, short-term prediction and adjustable evaluation are carried out.

[0083] Both shiftable loads and transferable loads transfer the time of power load from peak hours to off-peak hours without affecting the normal production and life of users. During the dispatch time, the total power consumption of such loads does not change, only the power consumption time changes. The difference between shiftable loads and transferable loads is whether the load can be fully transferred.

[0084] In one embodiment, in a smart grid system, a resource scheduling method is implemented to effectively manage and schedule resources including distributed power sources and uncertain loads. First, the system subdivides the uncertain loads into four categories: non-adjustable loads, curtailable loads, transferable loads, and shiftable loads. Through smart meters and load monitoring systems, the power consumption of these uncertain loads is obtained in real time to generate corresponding power consumption curves. For example, non-adjustable loads may include some key equipment whose power demand is fixed and cannot be adjusted; curtailable loads may be the power consumption of some non-key equipment, which can be reduced when necessary; transferable loads refer to power consumption that can be adjusted in time, such as shifting some peak hours to off-peak hours; shiftable loads refer to power demand that can be shifted between different dates.

[0085] Next, the system analyzes historical data and current power consumption patterns to determine the proportional relationship between loads that can be cut, loads that can be transferred, and loads that can be shifted. For example, it is found that the load that can be cut accounts for 20% of the total load, the load that can be transferred accounts for 30%, and the load that can be shifted accounts for 15%. Based on these proportional relationships, and combined with the preset maximum power consumption threshold and minimum power consumption threshold, the system calculates the adjustable space of the uncertain load. For example, if the maximum power consumption threshold is 100MW and the minimum power consumption threshold is 60MW, then the adjustable space is 40MW. In this way, the system can flexibly adjust the uncertain load while ensuring the stable operation of the power grid to adapt to the fluctuations and changes of distributed power sources.

[0086] In one embodiment, in a microgrid community, in order to improve energy efficiency and user satisfaction, a scheduling strategy that comprehensively considers distributed power sources and uncertain loads is implemented. First, the strategy divides uncertain loads into four categories: non-adjustable loads (such as basic living electricity such as refrigerators and air conditioners), curtailable loads (such as non-essential lighting and entertainment equipment electricity), transferable loads (such as electric vehicle charging, which can adjust the charging time) and shiftable loads (such as certain seasonal electricity, such as holiday decorations). By installing smart home devices and energy management systems, the power consumption of these loads is monitored and recorded in real time, generating detailed power consumption curves.

[0087] Then, based on the historical electricity consumption data and user behavior pattern analysis, the proportional relationship between various adjustable loads is determined. For example, it is found that the load that can be reduced accounts for 15% of the total community load, the load that can be transferred accounts for 25%, and the load that can be shifted accounts for 10%. Next, combined with the maximum carrying capacity of the community power grid (maximum power consumption threshold) and the minimum stable operation requirement (minimum power consumption threshold), the adjustable space of the uncertain load is calculated. For example, if the maximum power consumption threshold is 500kW and the minimum power consumption threshold is 300kW, then the adjustable space is 200kW. In this way, the community energy management system can effectively utilize the adjustable space while ensuring the reliability and quality of power supply, optimize the distribution and use of distributed power sources, reduce energy waste, and improve the energy efficiency of the entire community.

[0088] In one embodiment of the present application, the uncertain adjustable resource also includes an electric vehicle charging station. The server obtains historical data of the electric vehicle charging station and determines the scheduling potential of the electric vehicle charging station based on the historical data. The historical data is expressed by the following formula (1):

[0089] (1)

[0090] It should be noted that the embodiments of the present application represents the historical data of the electric vehicle charging station, Indicates electric vehicle Arrival time, Indicates the electric vehicle The departure time, Indicates the electric vehicle The battery level reaches Indicates the electric vehicle The battery power is left, Indicates the electric vehicle The upper limit of battery power, Indicates the electric vehicle The battery power limit is Indicates the electric vehicle The upper limit of charging power, Indicates the electric vehicle The lower limit of charging power.

[0091] In one embodiment, in a smart city, an advanced resource scheduling method is implemented to more effectively manage and schedule uncertain adjustable resources including electric vehicle charging stations. The method first considers electric vehicle charging stations as part of uncertain adjustable resources because the charging behavior of electric vehicles has greater flexibility and uncertainty.

[0092] In order to accurately evaluate the dispatching potential of electric vehicle charging stations, the system first obtains the historical data of the charging station. These data are expressed by specific formulas, which include key information such as the arrival time, departure time, arrival battery level, departure battery level, battery level upper limit, battery level lower limit, charging power upper limit and charging power lower limit of electric vehicles. This information fully reflects the behavior patterns and charging needs of electric vehicles at charging stations.

[0093] Through in-depth analysis of historical data, the system can determine the charging demand and potential of electric vehicle charging stations in different time periods. For example, the system finds that the charging demand of electric vehicles is low during certain time periods, while the charging power limit of the charging station is far from being fully utilized, which means that the charging station has greater scheduling potential during these time periods.

[0094] Based on this finding, the system can formulate a more flexible scheduling strategy. During the period of low charging demand, the system can guide electric vehicles to charge to make full use of the remaining charging capacity of the charging station; during the period of peak charging demand, the system can adjust the charging power to ensure the stable operation of the power grid and the charging needs of electric vehicles are met. In this way, the system not only improves the utilization rate of electric vehicle charging stations, but also effectively balances the load of the power grid, achieving efficient scheduling and optimal allocation of resources.

[0095] In one embodiment of the present application, the deterministic adjustable resources include an energy storage power station and a gas turbine. The server monitors the energy storage power station and obtains the charging and discharging status of the energy storage power station to determine the adjustable range of the energy storage power station; and monitors the gas turbine and obtains the operating status of the gas turbine to determine the adjustable range of the gas turbine.

[0096] In one embodiment, in an energy management system, the following strategy is implemented to more effectively schedule and utilize deterministic adjustable resources. These deterministic adjustable resources mainly include energy storage power stations and gas turbines. First, the system monitors the energy storage power station in real time and obtains the charging and discharging status of the energy storage power station through sensors and data acquisition equipment. These data include key indicators such as the current charging or discharging power, remaining capacity, and charging and discharging efficiency of the energy storage power station. By analyzing these data, the system can accurately determine the adjustable range of the energy storage power station in the current state, that is, the amount of power it can increase or decrease, and the sustainable time.

[0097] Secondly, the system also monitors the gas turbine in real time and obtains its operating status data. This data includes the current output power, fuel consumption rate, operating temperature, and emission level of the gas turbine. By analyzing these data, the system can evaluate the adjustable range of the gas turbine, that is, its power regulation ability and response speed under different operating conditions.

[0098] Based on the above monitoring and analysis of energy storage power stations and gas turbines, the energy management system can formulate more accurate and efficient scheduling strategies. For example, during peak hours of grid demand, the system can prioritize gas turbines to increase output power to meet the immediate needs of the grid; during low demand periods, the system can schedule energy storage power stations for charging to store excess electricity and release it when needed. In this way, the system achieves efficient utilization and optimal configuration of deterministic adjustable resources, improving the overall operating efficiency and reliability of the energy system.

[0099] 102. Based on the deep learning network, a model for output prediction of uncertain adjustable resources is constructed to determine the output prediction results corresponding to the uncertain adjustable resources.

[0100] A deep neural network (DNN) is a typical model in deep learning. It contains multiple hidden layers, so it is also called a multilayer perceptron. The perceptron itself is a two-layer structure model with multiple inputs and one output. A deep neural network is formed by improving and expanding the perceptron, such as adding multiple hidden layers to enhance the expressive power of the model and increasing the number of outputs to improve the functions of classification, regression, noise reduction, and clustering.

[0101] Figure 2 A schematic diagram of the deep neural network structure provided in the embodiment of the present application. Figure 2 As shown in the figure, the layers of a deep neural network are divided into three categories according to their type: input layer, hidden layer, and output layer. Each layer consists of several neurons and is connected in a fully connected form, and the neurons between layers are connected to each other. Each layer is connected with weights so that it contains the corresponding activation function and loss function, as well as the corresponding bias, which is used to evaluate the difference between the network output and the true output.

[0102] Specifically, the server constructs an uncertainty adjustable resource output prediction model based on a deep learning network and obtains sample data. In order to eliminate the dimensional differences between different types of variables and ensure the convergence of deep neural network training, the sample data is standardized; the processed sample data is input into the uncertainty adjustable resource output prediction model to encode the sample data through an activation function to obtain the corresponding hidden layer variables; the hidden layer variables are shown as follows (2)-(4):

[0103] (2)

[0104] (3)

[0105] (4)

[0106] It should be noted that the embodiments of the present application Represents the input matrix of processed sample data, represents the parameter matrix after being encoded by the activation function, Represents the total number of inputs, represents the linear relationship coefficient, Represents the error bias between the parameter matrix and the input matrix.

[0107] In one embodiment, in a smart grid system, in order to accurately predict the output of uncertain adjustable resources (such as wind farms, solar power stations, etc.), an output prediction model is constructed based on deep learning technology. The following are the specific implementation steps of the model: First, a large amount of historical sample data is collected from the smart grid system, which includes output data of uncertain adjustable resources, meteorological data (such as wind speed, light intensity, etc.), time data, etc. Then, the collected sample data is standardized to eliminate the impact of different dimensions and numerical ranges on model training. Standardization processing includes data normalization, denoising, filling missing values, etc.

[0108] Based on deep learning technology, an uncertainty adjustable resource output prediction model is constructed. The model adopts a multi-layer neural network structure, including an input layer, multiple hidden layers and an output layer. The processed sample data is input into the prediction model. In the input layer, the data is encoded by the activation function to obtain the corresponding hidden layer variables. These hidden layer variables can capture the complex nonlinear relationship between the input data, and by continuously adjusting these parameters, the model can learn the mapping relationship between the input data and the output.

[0109] The model is trained using optimization algorithms (such as gradient descent) to minimize the prediction error by iteratively updating parameters. During the training process, methods such as cross-validation are used to evaluate the generalization ability of the model to ensure that the model can maintain good prediction performance on different data sets. After the training is completed, the real-time collected uncertainty adjustable resource data and meteorological data are input into the prediction model to obtain the corresponding output prediction results.

[0110] In one embodiment of the present application, a deep neural network with multiple hidden layers has stronger feature extraction capabilities than a shallow network. The following formulas (5)-(7) represent the output of multiple hidden layers: And the decoded function The output layer The mathematical relationship between:

[0111] (5)

[0112] (6)

[0113] (7)

[0114] It should be noted that the embodiments of the present application Represents the output matrix after multiple hidden layers, Indicates that the decoded function The output layer parameter matrix, represents the linear relationship coefficient between the two layers, Represents the error bias between the output layer parameter matrix and the output quantity matrix.

[0115] In one embodiment, in a deep learning prediction model, in order to accurately predict the output of a certain uncertain adjustable resource (such as a wind power station), a neural network structure including multiple hidden layers is designed, and the output of the hidden layer is converted into the final prediction result through a specific decoding function. The following are the specific implementation steps of this process:

[0116] First, historical output data and meteorological data of wind power stations are collected and necessary preprocessing is performed, such as normalization and filling missing values. The processed data is used as an input matrix and input to the input layer of the deep learning model. The input data is nonlinearly transformed through multiple hidden layers, each of which contains an activation function to capture the complex relationship between the data. After being processed by multiple hidden layers, the input data is converted into an output matrix, which contains high-level features extracted from the original data.

[0117] The output matrix is ​​then fed into a decoding function that converts the output of the hidden layer into the final prediction result. The decoding function contains the output layer parameter matrix, which is connected to the output matrix through two layers of linear relationship coefficients and error bias. The model is trained using historical data and the model parameters are adjusted through an optimization algorithm to minimize the prediction error. After training, real-time data is fed into the model to obtain the output prediction result of the wind power station.

[0118] 103. Construct an evaluation strategy for the resource adjustable capacity corresponding to the multi-energy coupling virtual power plant, and determine the adjustable capacity corresponding to the multi-energy coupling virtual power plant by combining the output prediction results corresponding to the uncertain adjustable resources and the deterministic adjustable resources.

[0119] The resource adjustable capacity evaluation strategy for building a multi-energy coupled virtual power plant includes offline training of DNN models, online prediction of uncertain resource output data, and real-time evaluation of adjustable capacity.

[0120] When training a DNN model offline, in the training framework of the DNN model, the weight matrix is ​​usually used for standardized preprocessing to eliminate the dimensional differences between variables of different types, thereby ensuring the convergence of the DNN training results. The traditional method is to use the stochastic gradient descent method to update the weight matrix, but this method is time-consuming and has low accuracy. The Adam algorithm is an improved stochastic gradient descent method. As an adaptive learning optimizer, it further optimizes the weight matrix by adding dynamic vectors, thereby improving the calculation accuracy of the DNN model. Although this method adds some additional calculations, it improves efficiency and calculation accuracy compared to traditional methods.

[0121] Figure 3 Schematic diagram of the offline training framework of the DNN model based on the Adam optimizer improvement provided in the embodiment of the present application. Figure 3 As shown in the figure, first, the expected disturbance is used to solve the data, and the optimal diversity data sample database is constructed. The input samples and output samples are normalized and preprocessed. The samples after preprocessing are divided into training sample sets and test sample sets. After that, the number of hidden layers and the number of neurons in each hidden layer in the DNN model are set, and the sample data in the training sample set is input into the DNN for training. The input weight matrix and the hidden layer feature vector are solved layer by layer, and the weight matrix is ​​optimized by the Adam algorithm optimizer, and then the trained DNN model is tested using the sample data in the test sample set.

[0122] Specifically, in one embodiment of the present application, the server calculates the down-regulation range corresponding to the distributed power source by using the following formula (8):

[0123] (8)

[0124] It should be noted that the embodiments of the present application represents the downward adjustment range of the distributed power source, Indicates that the distributed power source is The predicted output at the time.

[0125] The server calculates the output operation down-adjustment boundary and output operation up-adjustment boundary of the electric vehicle charging station through the following formulas (9)-(10):

[0126] (9)

[0127] (10)

[0128] It should be noted that the embodiments of the present application Indicates that the electric vehicle charging station The lower limit of the moment, Indicates that the electric vehicle charging station The upward adjustment limit of the moment, represents the forecast data of electric vehicle charging stations, represents the lower bound of the operating capacity of the electric vehicle charging station, represents the upper bound of the operating capacity of the electric vehicle charging station.

[0129] The server calculates the adjustable capacity of uncertain load through the following formulas (11)-(12):

[0130] (11)

[0131] (12)

[0132] It should be noted that the embodiments of the present application Indicates that the uncertainty load is The load at the moment reduces the power. Indicates that the uncertainty load is The load at the moment increases the power. Indicated in The power reduction that can be achieved by reducing the load at any time, Indicated in The variable power of the load that can be transferred at any time, Indicated in The load can be shifted at any time. Indicated in The maximum output of the total load at any time, Indicated in The current output of the load at that moment.

[0133] In one embodiment, in a smart grid system, in order to effectively manage and dispatch uncertain loads, including curtailable loads, transferable loads, and shiftable loads, an advanced load dispatching strategy is implemented. The strategy first determines the proportional relationship between the three types of loads, and further determines the adjustable space corresponding to the uncertain load based on the preset maximum power consumption threshold and minimum power consumption threshold. The following are the specific implementation steps and calculation process:

[0134] First, the downward adjustment range of distributed power generation is calculated by a specific formula (8). This range is determined based on the output size of distributed power generation at the predicted time. Then, the output operation downward adjustment boundary and upward adjustment boundary of the electric vehicle charging station are calculated by specific formulas (9)-(10). These two boundaries are determined based on the predicted data of the electric vehicle charging station, the lower limit and upper limit of the operating capacity. Finally, the adjustable capacity of the uncertain load is also calculated by specific formulas (11)-(12). This capacity is determined based on the reduction power of the load that can be reduced, the change power of the load that can be transferred, the change power of the load that can be translated, the maximum output of the overall load, and the current output.

[0135] Based on the above calculations, a load dispatching strategy was developed. During the peak period of grid demand, priority is given to dispatching the load that can be cut, and the power consumption time of the transferable load and the shiftable load is adjusted to reduce the load pressure of the grid. During the trough period of grid demand, distributed power sources and electric vehicle charging stations can be dispatched for charging or energy storage in case of emergency. The formulated load dispatching strategy is implemented in the smart grid system and monitored and evaluated in real time. By comparing with the actual power consumption data, it is found that the strategy can effectively balance the load demand of the grid, improve the utilization efficiency of power resources, and reduce the cost of electricity.

[0136] Specifically, in one embodiment of the present application, the server calculates the adjustable power of the energy storage device through the following formulas (13)-(14):

[0137] (13)

[0138] (14)

[0139] It should be noted that the embodiments of the present application and Indicated in The adjustable capacity of the energy storage device at the time, Indicated in The SOC state of the energy storage device at the moment, represents the lower bound of the capacity of the energy storage device, represents the upper limit of the capacity of the energy storage device;

[0140] The adjustable power of the gas turbine is calculated by the following formulas (15)-(16):

[0141] (15)

[0142] (16)

[0143] It should be noted that the embodiments of the present application and Indicated in The adjustable power of the gas turbine at the moment, Indicated in The power generation capacity of the gas turbine at the time, represents the lower limit of the gas turbine output, represents the upper limit of the gas turbine output.

[0144] In one embodiment, in a multi-energy coupled virtual power plant system, in order to accurately evaluate its resource adjustable capacity, we constructed a set of evaluation strategies and combined the output prediction results of uncertain adjustable resources with deterministic adjustable resources to determine the adjustable capacity of the virtual power plant. The following are the specific implementation steps and calculation process:

[0145] First, the adjustable power of the energy storage device is calculated using specific formulas (13)-(14). This power is determined based on the adjustable capacity of the energy storage device at the prediction time, the SOC state (i.e., the state of charge of the energy storage device), the lower capacity limit, and the upper capacity limit. Next, the adjustable power of the gas turbine is calculated using specific formulas (15)-(16). This power is determined based on the adjustable power, power generation, lower output limit, and upper output limit of the gas turbine at the prediction time.

[0146] After calculating the adjustable power of energy storage equipment and gas turbines, these results are combined with the output forecast results of other deterministic adjustable resources (such as other types of generators) and uncertain adjustable resources (such as wind farms and solar power plants). By comprehensively considering the adjustable capacity and predicted output of all resources, the adjustable capacity of the multi-energy coupling virtual power plant at the forecast time can be determined. This adjustable capacity reflects the flexibility and response speed of the virtual power plant in meeting changes in grid demand.

[0147] The constructed resource adjustable capacity evaluation strategy is applied to the actual operation of multi-energy coupled virtual power plants, and real-time monitoring and optimization are carried out. By comparing and analyzing with actual operation data, the parameters and formulas in the evaluation strategy can be continuously adjusted and optimized to improve the accuracy and practicality of the adjustable capacity evaluation of virtual power plants.

[0148] The above is an embodiment of the method proposed in this application. Based on the same inventive concept, the embodiment of this application also provides an adjustable capacity evaluation device for a multi-energy coupled virtual power plant, the structure of which is as follows: Figure 4 shown.

[0149] Figure 4 This is a schematic diagram of the internal structure of an adjustable capacity evaluation device for a multi-energy coupled virtual power plant provided in an embodiment of the present application. Figure 4 As shown, the device includes:

[0150] at least one processor;

[0151] and, a memory communicatively coupled to the at least one processor;

[0152] The memory stores instructions that can be executed by at least one processor, and the instructions are executed by at least one processor to enable the at least one processor to:

[0153] Construct a multi-energy coupling resource model in a multi-energy coupling virtual power plant to determine the resource types in the multi-energy coupling virtual power plant; the resource types include uncertain adjustable resources and deterministic adjustable resources;

[0154] Based on the deep learning network, an output prediction model for uncertain adjustable resources is constructed to determine the output prediction results corresponding to uncertain adjustable resources;

[0155] A resource adjustable capacity evaluation strategy corresponding to the multi-energy coupling virtual power plant is constructed, and the adjustable capacity corresponding to the multi-energy coupling virtual power plant is determined by combining the output prediction results corresponding to the uncertain adjustable resources and the deterministic adjustable resources.

[0156] The present application also provides a non-volatile computer storage medium storing computer executable instructions. When the computer executable instructions are executed, they can:

[0157] Construct a multi-energy coupling resource model in a multi-energy coupling virtual power plant to determine the resource types in the multi-energy coupling virtual power plant; the resource types include uncertain adjustable resources and deterministic adjustable resources;

[0158] Based on the deep learning network, an output prediction model for uncertain adjustable resources is constructed to determine the output prediction results corresponding to uncertain adjustable resources;

[0159] A resource adjustable capacity evaluation strategy corresponding to the multi-energy coupling virtual power plant is constructed, and the adjustable capacity corresponding to the multi-energy coupling virtual power plant is determined by combining the output prediction results corresponding to the uncertain adjustable resources and the deterministic adjustable resources.

[0160] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0161] The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0162] The devices and media provided in the embodiments of the present application correspond one-to-one to the methods. Therefore, the devices and media also have similar beneficial technical effects as the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0163] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0164] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0165] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0166] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0167] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0168] Memory may include non-permanent storage in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0169] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0170] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0171] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for evaluating the adjustable capacity of a multi-energy coupled virtual power plant, characterized in that: The method comprises: Constructing a multi-energy coupling resource model in a multi-energy coupling virtual power plant to determine the resource types in the multi-energy coupling virtual power plant; the resource types include uncertain adjustable resources and deterministic adjustable resources; Based on the deep learning network, an output prediction model of an uncertain adjustable resource is constructed to determine the output prediction result corresponding to the uncertain adjustable resource; Constructing a resource adjustable capacity evaluation strategy corresponding to the multi-energy coupling virtual power plant, and combining the output prediction results corresponding to the uncertain adjustable resources and the deterministic adjustable resources to determine the adjustable capacity corresponding to the multi-energy coupling virtual power plant; The uncertain adjustable resources include distributed power sources and uncertain loads; The method further comprises: The uncertain load is divided into non-adjustable load, curtailable load, transferable load and shiftable load, and the power consumption corresponding to the uncertain load is obtained to generate a corresponding power consumption curve; Determine the proportional relationship between the curtailable load, the transferable load and the shiftable load, and determine the adjustable space corresponding to the uncertain load based on a preset maximum power consumption threshold and a minimum power consumption threshold; The determining of the proportional relationship between the reducible load, the transferable load, and the shiftable load, and determining the adjustable space corresponding to the uncertain load based on a preset maximum power consumption threshold and a minimum power consumption threshold, specifically includes: The corresponding down-regulation range of distributed power generation is calculated by the following formula: S der (t)=P der (t) Among them, S der (t) represents the down-regulation range of the distributed power source, P der (t) represents the predicted output of the distributed power source at time t; The output operation downward adjustment boundary and output operation upward adjustment boundary of the electric vehicle charging station are calculated by the following formula: S ev,ch (t)=P ev (t)-P ev,min S ev,dis (t)=P ev,max -P ev (t) Among them, S ev,ch (t) represents the downward adjustment boundary of the electric vehicle charging station at time t, S ev,dis (t) represents the upper adjustment limit of the electric vehicle charging station at time t, P ev (t) represents the forecast data of electric vehicle charging stations, P ev,min represents the lower bound of the operating capacity of the electric vehicle charging station, P ev,max represents the upper limit of the operating capacity of the electric vehicle charging station; The adjustable capacity of uncertain load is calculated by the following formula: S load,ch (t)=ΔP cut (t)+ΔP trans (t)+ΔP change (t) S load,dis (t)=ΔP load,max (t)-ΔP load (t)+ΔP trans (t)+ΔP change (t) Among them, S losd,ch (t) represents the load reduction power of the uncertain load at time t, S load,dis (t) represents the load increase power of the uncertain load at time t, ΔP cut (t) represents the power reduction that can be achieved at time t, ΔP trans (t) represents the change in power of the transferable load at time t, ΔP change (t) represents the change in power of the load that can be translated at time t, P load,max (t) represents the maximum output of the overall load at time t, P load (t) represents the current output of the load at time t.

2. The method for evaluating the adjustable capacity of a multi-energy coupled virtual power plant according to claim 1, characterized in that: The uncertain adjustable resources also include electric vehicle charging stations; The method further comprises: Acquiring historical data of the electric vehicle charging station, and determining the dispatching potential of the electric vehicle charging station based on the historical data; The historical data is expressed by the following formula: Among them, D ev represents the historical data of the electric vehicle charging station, represents the arrival time of electric vehicle n, represents the departure time of the electric vehicle n, represents the reached battery capacity of the electric vehicle n, represents the leaving battery capacity of the electric vehicle n, represents the upper limit of the battery power of the electric vehicle n, represents the lower limit of the battery power of the electric vehicle n, represents the upper limit of the charging power of the electric vehicle n, represents the lower limit of the charging power of the electric vehicle n.

3. The method for evaluating the adjustable capacity of a multi-energy coupled virtual power plant according to claim 1, characterized in that: The deterministic adjustable resources include energy storage power stations and gas turbines; The method further comprises: Monitor the energy storage power station and obtain the charging and discharging status corresponding to the energy storage power station to determine the adjustable range corresponding to the energy storage power station; The gas turbine is monitored and the corresponding operating status of the gas turbine is obtained to determine the corresponding adjustable range of the gas turbine.

4. The method for evaluating the adjustable capacity of a multi-energy coupled virtual power plant according to claim 1, characterized in that: The method of constructing an output prediction model of an uncertain adjustable resource based on a deep learning network to determine an output prediction result corresponding to the uncertain adjustable resource specifically includes: Based on the deep learning network, an uncertain adjustable resource output prediction model is constructed, and sample data is obtained to perform standardization processing on the sample data; Inputting the processed sample data into the uncertainty adjustable resource output prediction model to encode the sample data through an activation function to obtain corresponding hidden layer variables; The hidden layer variables are as follows: <h2 style=";text-align:left;direction:ltr">Y = [y1, y2, y3...y<h2 style=";text-align:left;direction:ltr"> n <h2 style=";text-align:left;direction:ltr"> ]<h2 style=";text-align:left;direction:ltr"> T in, represents the input matrix of the processed sample data, Y represents the parameter matrix after encoding by the activation function, n represents the total number of inputs, and w i represents the linear relationship coefficient, b represents the error bias between the parameter matrix and the input matrix, x i represents the i-th sample data, Represents the activation function.

5. The method for evaluating the adjustable capacity of a multi-energy coupled virtual power plant according to claim 4, characterized in that: The method further comprises: The output after multiple hidden layers is expressed by the following formula: And after decoding function g θ The mathematical relationship between the output layer Z is: Z=[z1,z2,z3…z n ] in, represents the output matrix after multiple hidden layers, and Z represents the output matrix after the decoding function g θ The output layer parameter matrix, w i ′ Represents the linear relationship coefficient between the two layers, b ′ represents the error bias between the output layer parameter matrix and the output quantity matrix, represents the nth hidden layer variable.

6. The method for evaluating the adjustable capacity of a multi-energy coupling virtual power plant according to claim 3, characterized in that: The constructing of the resource adjustable capacity evaluation strategy corresponding to the multi-energy coupling virtual power plant and combining the output prediction results corresponding to the uncertain adjustable resources and the deterministic adjustable resources to determine the adjustable capacity corresponding to the multi-energy coupling virtual power plant specifically includes: The adjustable power of the energy storage device is calculated using the following formula: S es,ch (t)=SOC(t)-S es,min S es,dis (t)=S es,max -SOC(t) Among them, S es,ch (t) and S es,dis (t) represents the adjustable capacity of the energy storage device at time t, SOC(t) represents the SOC state of the energy storage device at time t, S es,min represents the lower limit of the capacity of the energy storage device, S es,max represents the upper limit of the capacity of the energy storage device; The adjustable power of the gas turbine is calculated by the following formula: S gs,ch (t)=P(t)-P gs,min S gs,dis (t)=P gs,max -P(t) Among them, S gs,ch (t) and S gs,dis P(t) represents the adjustable power of the gas turbine at time t, P(t) represents the power generation power of the gas turbine at time t, P gs,min represents the lower limit of the gas turbine output, P gs,max represents the upper limit of the gas turbine output.

7. An adjustable capacity evaluation device for a multi-energy coupled virtual power plant, characterized in that: The device comprises: at least one processor; and, a memory communicatively coupled to the at least one processor; Wherein, the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the adjustable capacity assessment method of a multi-energy coupled virtual power plant as described in any one of claims 1-6.

8. A non-volatile computer storage medium storing computer executable instructions, characterized in that: When the computer executable instructions are executed, the adjustable capacity evaluation method of a multi-energy coupled virtual power plant as described in any one of claims 1 to 6 is implemented.

Images