Electric heating coupling system scheduling method and scheduling system based on renewable energy sources

Abstract

The invention discloses an electrothermal coupling system scheduling method and scheduling system based on renewable energy sources. The method comprises the following steps: generating a graph structure which is matched with a to-be-scheduled electrothermal coupling system and comprises a plurality of nodes; determining an actual power value of each component based on the data of the to-be-scheduled electrothermal coupling system, and determining a power output value corresponding to each node by using a graph attention mechanism; for each node, determining a deviation score of the node based on the difference between the actual power value of the component corresponding to the node and the power output value corresponding to the node; determining a bottleneck node from the plurality of nodes based on the deviation score; and configuring the optimization model based on the bottleneck node, and scheduling the to-be-scheduled electrothermal coupling system according to the configured optimization model, so that the scheduling of the electrothermal coupling system can be optimized, and the scheduling efficiency and reliability of the electrothermal coupling system can be improved.

Description

Technical Field

[0001] The present application relates to the field of scheduling technology, and in particular to a scheduling method and a scheduling system for an electric-thermal coupling system based on renewable energy. Background Art

[0002] As the proportion of renewable energy grows rapidly, renewable energy is expected to become the main energy source in the future. However, the existing power system's insufficient ability to effectively absorb renewable energy has become a key bottleneck restricting the development of renewable energy.

[0003] At present, the electric-thermal coupling system has become an important solution to the bottleneck problem of renewable energy consumption due to its ability to flexibly allocate electric and thermal energy resources. The electric-thermal coupling system can alleviate the pressure of power consumption through heat storage technology and provide a stable energy utilization path on the demand side.

[0004] However, the system composition of the electrothermal coupling system is very complex, involving energy flows in multiple dimensions (such as using a certain type of energy to generate electricity, or converting excess electricity into storable thermal energy). Existing scheduling methods for electrothermal coupling systems usually focus on scheduling the system by considering the energy flow conditions of a single dimension separately. It is difficult to effectively deal with the problems caused by the complex dynamic behavior and uncertainty factors in the electrothermal coupling system when renewable energy is connected to the grid, resulting in low scheduling efficiency and reliability. Summary of the invention

[0005] In order to solve the above technical problems, the embodiments of the present application propose a scheduling method and a scheduling system for an electric-thermal coupling system based on renewable energy, which can improve the efficiency and reliability of the scheduling of the electric-thermal coupling system.

[0006] In a first aspect, an embodiment of the present application provides a method for scheduling an electric-thermal coupling system based on renewable energy, comprising: Generate a graph structure that matches the electric-thermal coupling system to be scheduled and includes a plurality of nodes, wherein the electric-thermal coupling system to be scheduled includes a plurality of components, the components are used to indicate renewable energy consumption information in the electric-thermal coupling system to be scheduled, and the plurality of nodes correspond to the plurality of components one by one; Based on the data of the electric-thermal coupling system to be scheduled, determining the actual power value of each component, and determining the power output value corresponding to each node by using a graph attention mechanism; For each node, determine a deviation score for the node based on a difference between an actual power value of a component corresponding to the node and a power output value corresponding to the node; determining a bottleneck node from the plurality of nodes based on the deviation score; Based on the bottleneck node, an optimization model is configured to schedule the electric-thermal coupling system to be scheduled according to the configured optimization model.

[0007] Optionally, generating a graph structure matching the electric-thermal coupling system to be scheduled and including a plurality of nodes includes: Model each component to generate corresponding nodes; Based on the data of each component contained in the data of the electric-thermal coupling system to be scheduled, using the node embedding technology to calculate the embedding vector of the node corresponding to each component; Based on the multiple nodes and their respective embedding vectors, an adjacency matrix of the graph structure is constructed by calculating an optimal embedding vector, thereby obtaining the graph structure.

[0008] Optionally, based on the data of the electric-thermal coupling system to be scheduled, a power output value corresponding to each node is determined using a graph attention mechanism, including: Extracting input features of nodes corresponding to each component from data of each component included in the data of the electric-thermal coupling system to be scheduled; Determining attention weights based on input features of each of the plurality of nodes; Based on the attention weight, the input features of each of the multiple nodes and the weight matrix corresponding to the graph structure, using the activation function in the graph attention mechanism, calculate and obtain the aggregate representation corresponding to each node; Based on the aggregate representation corresponding to each node, the graph attention mechanism is used to determine the power output value corresponding to each node.

[0009] Optionally, the determining the power output value corresponding to each node using a graph attention mechanism based on the aggregate representation corresponding to each node includes: The aggregation corresponding to each node is represented as the power of each node at time t, and the power of each node at time t is input into the graph attention mechanism for iterative prediction, so that the graph attention mechanism outputs the power of each node at time t+k obtained through iterative prediction as the power output value, where t is a positive real number and k is a positive real number.

[0010] Optionally, the deviation score is calculated by the following formula: in, is the actual power value of the component corresponding to node i at time t, is the power output value corresponding to node i, is the standard deviation of the output value corresponding to node i, is the deviation score of node i at time t.

[0011] Optionally, the bottleneck nodes are a given number of nodes corresponding to the highest deviation scores among the multiple nodes, wherein the bottleneck nodes are used to reflect key constraints in the renewable energy consumption information.

[0012] Optionally, the configured optimization model includes an objective function and constraints; The objective function is used to indicate the minimization of the total cost, which includes the power generation cost, the wind abandonment cost and the state variable cost of the bottleneck node, and the state variable cost is calculated by the state variable and the state penalty coefficient of the bottleneck node.

[0013] Optionally, the constraint condition includes at least one of the following: power balance constraint, thermal balance constraint, bottleneck node status constraint; The bottleneck node state constraint is used to limit the state variable of the bottleneck node to be within a set safety threshold range.

[0014] In a second aspect, an embodiment of the present application provides a scheduling system, including: A graph structure generating module, used to generate a graph structure matching the electric-thermal coupling system to be scheduled and including a plurality of nodes, wherein the electric-thermal coupling system to be scheduled includes a plurality of components, the components are used to indicate renewable energy consumption information in the electric-thermal coupling system to be scheduled, and the plurality of nodes correspond to the plurality of components one by one; A power determination module, used to determine the actual power value of each component based on the data of the electric-thermal coupling system to be scheduled, and to determine the power output value corresponding to each node by using a graph attention mechanism; A deviation scoring module is used to determine, for each node, a deviation score of the node based on a difference between an actual power value of a component corresponding to the node and a power output value corresponding to the node; A bottleneck node determination module, configured to determine a bottleneck node from the plurality of nodes based on the deviation score; The optimization scheduling module is used to configure the optimization model based on the bottleneck node, so as to schedule the electric-thermal coupling system to be scheduled according to the configured optimization model.

[0015] Optionally, based on the data of the electric-thermal coupling system to be scheduled, a power output value corresponding to each node is determined using a graph attention mechanism, including: Extracting input features of nodes corresponding to each component from data of each component included in the data of the electric-thermal coupling system to be scheduled; Determining attention weights based on input features of each of the plurality of nodes; Based on the attention weight, the input features of each of the multiple nodes and the weight matrix corresponding to the graph structure, using the activation function in the graph attention mechanism, calculate and obtain the aggregate representation corresponding to each node; Based on the aggregate representation corresponding to each node, the graph attention mechanism is used to determine the power output value corresponding to each node.

[0016] In summary, the embodiments of the present application have at least the following beneficial effects: According to an embodiment of the present application, a graph structure is generated that matches the electric-thermal coupling system to be scheduled and includes multiple nodes, wherein the electric-thermal coupling system to be scheduled includes multiple components, and the components are used to indicate the renewable energy consumption information in the electric-thermal coupling system to be scheduled, and the multiple nodes correspond to the multiple components one by one; based on the data of the electric-thermal coupling system to be scheduled, the actual power value of each component is determined, and the power output value corresponding to each node is determined by using the graph attention mechanism; for each node, the deviation score of the node is determined based on the difference between the actual power value of the component corresponding to the node and the power output value corresponding to the node; a bottleneck node is determined from the multiple nodes based on the deviation score; based on the bottleneck node, an optimization model is configured to schedule the electric-thermal coupling system to be scheduled according to the configured optimization model, so as to optimize the scheduling of the electric-thermal coupling system and improve the efficiency and reliability of the scheduling of the electric-thermal coupling system. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flow chart of a method for dispatching an electric-thermal coupling system based on renewable energy provided in an embodiment of the present application; Figure 2 It is a schematic diagram of the abnormality scoring trend in the model training, verification and testing stages provided in the embodiment of the present application; Figure 3 It is a schematic diagram comparing the model prediction value and the actual target value provided in the embodiment of the present application; Figure 4 is a schematic diagram of a bottleneck node provided in an embodiment of the present application; Figure 5 It is a structural diagram of the scheduling system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0019] In the description of the present application, the terms "first", "second", "third", etc. are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, features defined as "first", "second", "third", etc. may explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise specified, the meaning of "multiple" is two or more. In the description of the present application, the term "including" and its variations are open inclusions, i.e., "including but not limited to". The term "based on" means "at least partially based on". The term "according to" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments".

[0020] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0021] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this application have the same meanings as those commonly understood by those skilled in the art. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood by specific circumstances.

[0022] First, see Figure 1 , shows a schematic flow chart of a method for dispatching an electric-thermal coupling system based on renewable energy provided in an embodiment of the present application, the method comprising steps S101-S105, which are specifically as follows: S101, generating a graph structure that matches the electric-thermal coupling system to be scheduled and includes a plurality of nodes, wherein the electric-thermal coupling system to be scheduled includes a plurality of components, and the components are used to indicate renewable energy consumption information in the electric-thermal coupling system to be scheduled, and the plurality of nodes correspond one-to-one to the plurality of components.

[0023] In some cases, the multiple components in the electric-thermal coupling system to be scheduled may include a traditional condensing power generation unit, a cogeneration unit, a wind power unit and an electric boiler unit, which are respectively suitable for reflecting the status information of the corresponding equipment in the electric-thermal coupling system to be scheduled, wherein the energy balance equation that the electric-thermal coupling system to be scheduled needs to satisfy can be constructed in the following manner: obtaining the data of the traditional condensing power generation unit, the data of the cogeneration unit, the data of the wind power unit and the data of the electric boiler unit; constructing an energy balance equation based on the data of the traditional condensing power generation unit, the data of the cogeneration unit, the data of the wind power unit and the data of the electric boiler unit, wherein the energy balance equation is suitable for characterizing the temporal and spatial dependence of the output of electric energy and thermal energy in the model.

[0024] In some specific examples, the energy balance equation can be determined by the following formula: in, represents the output power of the traditional condensing power generation unit at time t, represents the output power of the CHP unit at time t, represents the wind power unit output power at time t, represents the power consumption of the electric boiler at time t, represents the system load power demand at time t, represents the transmission loss at time t.

[0025] In this embodiment, the cogeneration unit is capable of generating both electrical energy and thermal energy, and adjusting its electrical output according to thermal demand. The wind power generation unit mainly relies on natural wind power, and its output is intermittent and unstable, while the electric boiler consumes electrical energy to generate thermal energy to supplement thermal demand. In order to better utilize the synergy of these devices, they must be accurately modeled and coordinated. In order to ensure the stability of the system under different operating conditions, an energy balance equation can be constructed to ensure that the output of electricity and heat matches the demand. In particular, when wind power generation fluctuates greatly, how to maintain the balance of the system by adjusting the power output of the cogeneration unit and the power consumption of the electric boiler.

[0026] Furthermore, considering the uncertainty of wind power output and the different working modes of cogeneration units in different seasons, the output of electricity and heat can be dynamically analyzed in time series. By modeling the changes in power load, heat demand and wind power output in different time periods, the system can be optimized for scheduling throughout the year. In addition, in order to ensure the stable operation of the system, multi-scenario simulation tests can be carried out, especially when wind power generation fluctuates greatly, to simulate scenarios with different power demand and heat demand, and analyze the energy balance capacity of each component. Through this simulation, the extreme situations that may occur in actual operation are determined, and a basis is provided for subsequent optimization.

[0027] S102, based on the data of the electric-thermal coupling system to be scheduled, determining the actual power value of each component, and determining the power output value corresponding to each node by using a graph attention mechanism.

[0028] It should be noted that the graph attention mechanism is a variant of GNN (Graph Neural Networks), which introduces an attention mechanism to improve the aggregation process of node features. Compared with traditional GNN, the graph attention mechanism does not treat the information of all neighbor nodes equally, but dynamically assigns different weights according to the importance of neighbor nodes. This mechanism allows the model to more flexibly focus on neighbor information that is more important to the current task, thereby improving the performance of the model.

[0029] S103 , for each node, determining a deviation score of the node based on a difference between an actual power value of a component corresponding to the node and a power output value corresponding to the node.

[0030] In one example, the difference may be characterized by calculating the difference between the actual power value and the power output value.

[0031] In one example, determining the deviation score of the node includes: determining the standard deviation of the output value corresponding to the node; calculating the difference between the actual power value and the power output value; calculating the absolute value of the difference and the ratio between the standard deviation of the output value corresponding to the node; and determining the deviation score of the node based on the ratio.

[0032] S104: Determine a bottleneck node from the multiple nodes based on the deviation score.

[0033] S105, configuring an optimization model based on the bottleneck node, so as to schedule the electric-thermal coupling system to be scheduled according to the configured optimization model.

[0034] In one example, the optimization model to be configured can be a model representing the optimal solution determined by linear programming, nonlinear programming, dynamic programming, integer programming and / or system science methods. Here, the optimization model can be configured with some objectives and / or constraints associated with the state variables of the bottleneck node to complete the configuration.

[0035] In an optional implementation, generating a graph structure matching the electric-thermal coupling system to be scheduled and including a plurality of nodes includes: Each component is modeled to generate corresponding nodes.

[0036] Based on the data of each component contained in the data of the electric-thermal coupling system to be scheduled, an embedding vector of a node corresponding to each component is calculated using a node embedding technology.

[0037] Based on the multiple nodes and their respective embedding vectors, an adjacency matrix of the graph structure is constructed by calculating an optimal embedding vector, thereby obtaining the graph structure.

[0038] It should be noted that in this embodiment, each component in the system can be modeled as a node in a graph structure, and a feature vector of each node can be generated using node embedding technology, wherein the feature vector is suitable for characterizing the time series data, device type, and dynamic behavior corresponding to the node. In addition, the adjacency matrix is ​​constructed based on the interaction between the nodes. The graph structure is suitable for characterizing the mutual correlation between the nodes, and the correlation is suitable for reflecting the interdependence between the devices in the system. For example, the working state of the electric boiler may be directly affected by the output of the wind power generation unit. By constructing this graph structure, the energy flow path between the nodes can be clarified.

[0039] In some specific examples, the calculation formula for node embedding is as follows: in, is the embedding vector of node i, is the input feature of node i, is the embedded function.

[0040] In some specific examples, each element in the adjacency matrix A Suitable for characterizing nodes and nodes The correlation between them can be calculated by the following formula: in, Indicates that the top most relevant nodes.

[0041] In an optional implementation, based on the data of the electric-thermal coupling system to be scheduled, a power output value corresponding to each node is determined using a graph attention mechanism, including: From the data of each component contained in the data of the electric-thermal coupling system to be scheduled, the input features of the nodes corresponding to each component are extracted.

[0042] Attention weights are determined based on input features of each of the multiple nodes.

[0043] Based on the attention weight, the input features of each of the multiple nodes and the weight matrix corresponding to the graph structure, the activation function in the graph attention mechanism is used to calculate and obtain the aggregation representation corresponding to each node.

[0044] Based on the aggregate representation corresponding to each node, the graph attention mechanism is used to determine the power output value corresponding to each node.

[0045] In some specific examples, the aggregate representation can be calculated using the following formula: in, is the aggregate representation of node i, is the trainable weight matrix corresponding to the graph structure, is the attention weight, is the set of neighbor nodes of node i, represents the activation function, j represents A node in .

[0046] in, In this embodiment, in order to improve the system's adaptability to complex scenarios, this embodiment introduces a graph attention mechanism so that each node not only relies on its own historical behavior, but is also affected by its associated nodes. Through the graph attention mechanism, the power output of each node at a future moment can be predicted more accurately, thereby improving the overall scheduling efficiency of the system.

[0047] In an optional implementation, the determining the power output value corresponding to each node using a graph attention mechanism based on the aggregate representation corresponding to each node includes: The aggregation corresponding to each node is represented as the power of each node at time t, and the power of each node at time t is input into the graph attention mechanism for iterative prediction, so that the graph attention mechanism outputs the power of each node at time t+k obtained through iterative prediction as the power output value, where t is a positive real number and k is a positive real number.

[0048] In some specific examples, the graph attention mechanism can be used to perform multi-level predictions on each node in stages to achieve iterative prediction.

[0049] First, a short-term power output prediction is performed, that is, the power of each node at time t is input into the graph attention mechanism for the first prediction, which can be expressed as the following formula: in, Representation Node exist The short-term power prediction at the moment (i.e., the node predicted by the graph attention mechanism exist power at the moment), represents the prediction function of the graph attention mechanism, represents the power of node i at time t (i.e., the aggregate representation of node i at time t).

[0050] Then, the prediction results are extended to the medium and long-term trends. Through iterative prediction, the power output value of each node is gradually updated, which can be expressed as the following formula: in, Represents the predicted node exist The power of the moment, Indicates that node i is The power of the moment.

[0051] In this embodiment, through this step-by-step iterative prediction method, the model can capture the medium- and long-term power change trend of the node and form a multi-level prediction system. The phased prediction method in this embodiment can make the system more flexible to respond to instantaneous volatility changes, especially when the wind power generation output fluctuates violently, and can make scheduling adjustments in advance.

[0052] In an optional embodiment, the deviation score is calculated by the following formula: in, is the actual power value of the component corresponding to node i at time t, is the power output value corresponding to node i, is the standard deviation of the output value corresponding to node i, is the deviation score of node i at time t.

[0053] In this embodiment, nodes with higher deviation scores are nodes with larger output fluctuations. These nodes may be bottlenecks in the system, affecting the stable operation of the overall system. Especially when wind power output is unstable, some nodes may show obvious abnormal behavior and need further optimization. The deviation score in this embodiment is suitable for quantifying the degree of abnormality of each node. Each node is scored according to the deviation of its output value. Nodes with higher scores represent greater impact on the system. By sorting the deviation scores of all nodes, the bottleneck nodes in the system can be quickly located.

[0054] In addition, it should be understood that the bottleneck node of the system is not static and may shift with time and changes in operating conditions. Therefore, this embodiment not only analyzes the node behavior at a single moment, but also conducts a comprehensive evaluation of multiple moments through a dynamic detection method to ensure that the operating status of the system can be continuously monitored and new bottleneck nodes can be discovered in a timely manner.

[0055] In an optional implementation, the bottleneck nodes are a given number of nodes corresponding to the highest deviation scores among the multiple nodes, wherein the bottleneck nodes are used to reflect key constraints in the renewable energy consumption information.

[0056] In some cases, the bottleneck node may also be each node among multiple nodes whose corresponding deviation score is higher than the score threshold.

[0057] In an optional implementation, the configured optimization model includes an objective function and constraints.

[0058] The objective function is used to indicate the minimization of the total cost, which includes the power generation cost, the wind abandonment cost and the state variable cost of the bottleneck node, and the state variable cost is calculated by the state variable and the state penalty coefficient of the bottleneck node.

[0059] In some specific examples, the objective function can be expressed as the following formula: in, is the cost coefficient of condensing power generation, is the cost coefficient of cogeneration, is the cost coefficient of the wind power unit, represents the amount of wind power abandoned at time t, represents the output power of the traditional condensing power generation unit at time t, represents the output power of the CHP unit at time t, represents the wind power unit output power at time t, represents the power consumption of the electric boiler at time t, is the penalty coefficient for abandoned wind power, For Node (at this time is the state penalty coefficient of the bottleneck node at time t, For Node The state variable at time t may include voltage, current and / or temperature, and B is a set of bottleneck nodes.

[0060] In this embodiment, in order to reduce the bottleneck in the system and improve the absorption capacity of renewable energy, a multi-objective optimization model is designed. The main optimization objectives include reducing the amount of wind power abandoned, improving the operating efficiency of the system, and ensuring the stability of energy balance. During the optimization process, the system needs to consider multiple competing objectives, so it is necessary to balance the weights of each objective through a multi-objective optimization algorithm. During the optimization process, this embodiment dynamically adjusts the output power of each device by real-time monitoring of the operating status of wind power, cogeneration unit and electric boiler. Especially during periods of high wind power output, the power consumption of the electric boiler is increased to reduce the load on the power grid and ensure the energy balance of the system.

[0061] In addition, in order to further improve the dispatch flexibility of the system, this embodiment can also integrate distributed energy storage equipment into the electric-thermal coupling system. Through the distributed energy storage equipment, the excess electric energy of wind power can be stored and released during periods of weak wind or high heat demand, thereby alleviating the impact of wind power output fluctuations on the system.

[0062] In an optional implementation, the constraint condition includes at least one of the following: a power balance constraint, a thermal balance constraint, and a bottleneck node status constraint.

[0063] The bottleneck node state constraint is used to limit the state variable of the bottleneck node to be within a set safety threshold range.

[0064] In some specific examples, the power balance constraint can be determined by the following formula: in, represents the output power of the traditional condensing power generation unit at time t, represents the output power of the CHP unit at time t, represents the wind power unit output power at time t, represents the power consumption of the electric boiler at time t, represents the system load power demand at time t, represents the transmission loss at time t.

[0065] In some specific examples, the thermal balance constraint can be determined by the following formula: in, represents the heat generated by the cogeneration unit at time t, represents the heat generated by the electric boiler at time t, Represents the actual heat load demand at time t.

[0066] In some specific examples, the bottleneck node state constraint can be determined by the following formula: in, For Node The state variable at time t may include voltage, current and / or temperature. B is the bottleneck node set. Indicates the set safety threshold range corresponding to the state variable of the bottleneck node.

[0067] In this embodiment, the possibility of bottleneck nodes causing system operation risks can be reduced, and system reliability can be improved. In practical applications, the power demand and heat demand in different time periods may change, so the constraints in the optimization process need to be dynamically adjusted according to real-time data. For example, when the wind power output is low, it may be necessary to give priority to meeting the power demand, while in the period of high heat demand, it is necessary to give priority to meeting the heat supply. Through this dynamic adjustment, the system can better adapt to complex operating environments.

[0068] A specific example and its experimental results are given below to demonstrate the technical effect of the present application.

[0069] First, the data set of this example comes from the simulation results of the electric-thermal coupling system model, with a total of 696 records, each 24 records constitute a variable, a total of 30 variables and 12 channels, each channel represents a different heat loss coefficient simulation scenario. Among them, 144 records are used for the training set, 72 records are used for the validation set, and the remaining records are used for the test set. The following Table 1 shows the data division of the training set, validation set, and test set used in this embodiment. Before training, all variables were normalized using the upper limit of each variable as a reference value.

[0070] Table 1 Then, see Figure 2 , this example shows the anomaly score trends during the training, validation, and testing phases. Each phase is evaluated using an anomaly score (equivalent to a deviation score) with a preset threshold of 3.3971 to distinguish between normal and abnormal behavior. The scores during the training and validation phases are consistently low, indicating that the model is effective in learning and generalizing. However, a significant peak in the anomaly score appears around variables 310 to 361 during the testing phase, indicating that anomalies are detected.

[0071] Furthermore, see Figure 3 ,This embodiment compares the model prediction values ​​and the actual target values ​​in bottleneck ,detection, and shows the comparison between the normalized target value (solid line) and the model prediction ,value (dashed line) in 696 time steps. The results show the overall accuracy of the ,model, and the deviation can be clearly seen where the prediction value deviates from the ,target value.

[0072] Finally, see Figure 4 , this embodiment marks specific bottleneck nodes where the difference between the prediction and the target is significant. These points reflect the critical prediction deviations of the model at certain time steps (i.e., steps 301 to 330), indicating that these locations have become the main bottleneck factors for renewable energy power consumption.

[0073] On the second aspect, accordingly, the embodiments of the present application also provide a scheduling system that can implement all processes of the scheduling method of the electric-thermal coupling system based on renewable energy provided in the above embodiments.

[0074] See also Figure 5 , shows a schematic diagram of the structure of a scheduling system provided in an embodiment of the present application, the scheduling system comprising: The graph structure generation module 501 is used to generate a graph structure that matches the electric-thermal coupling system to be scheduled and includes multiple nodes, wherein the electric-thermal coupling system to be scheduled includes multiple components, and the components are used to indicate the renewable energy consumption information in the electric-thermal coupling system to be scheduled, and the multiple nodes correspond to the multiple components one by one.

[0075] The power determination module 502 is used to determine the actual power value of each component based on the data of the electric-thermal coupling system to be scheduled, and to determine the power output value corresponding to each node by using a graph attention mechanism.

[0076] The deviation scoring module 503 is used to determine, for each node, a deviation score of the node based on the difference between the actual power value of the component corresponding to the node and the power output value corresponding to the node.

[0077] The bottleneck node determination module 504 is configured to determine a bottleneck node from the plurality of nodes based on the deviation score.

[0078] The optimization scheduling module 505 is used to configure an optimization model based on the bottleneck node, so as to schedule the electric-thermal coupling system to be scheduled according to the configured optimization model.

[0079] In an optional implementation, generating a graph structure matching the electric-thermal coupling system to be scheduled and including a plurality of nodes includes: Each component is modeled to generate corresponding nodes.

[0080] Based on the data of each component contained in the data of the electric-thermal coupling system to be scheduled, an embedding vector of a node corresponding to each component is calculated using a node embedding technology.

[0081] Based on the multiple nodes and their respective embedding vectors, an adjacency matrix of the graph structure is constructed by calculating an optimal embedding vector, thereby obtaining the graph structure.

[0082] In an optional implementation, based on the data of the electric-thermal coupling system to be scheduled, a power output value corresponding to each node is determined using a graph attention mechanism, including: From the data of each component contained in the data of the electric-thermal coupling system to be scheduled, the input features of the nodes corresponding to each component are extracted.

[0083] Attention weights are determined based on input features of each of the multiple nodes.

[0084] Based on the attention weight, the input features of each of the multiple nodes and the weight matrix corresponding to the graph structure, the activation function in the graph attention mechanism is used to calculate and obtain the aggregation representation corresponding to each node.

[0085] Based on the aggregate representation corresponding to each node, the graph attention mechanism is used to determine the power output value corresponding to each node.

[0086] In an optional implementation, the determining the power output value corresponding to each node using a graph attention mechanism based on the aggregate representation corresponding to each node includes: The aggregation corresponding to each node is represented as the power of each node at time t, and the power of each node at time t is input into the graph attention mechanism for iterative prediction, so that the graph attention mechanism outputs the power of each node at time t+k obtained through iterative prediction as the power output value, where t is a positive real number and k is a positive real number.

[0087] In an optional embodiment, the deviation score is calculated by the following formula: in, is the actual power value of the component corresponding to node i at time t, is the power output value corresponding to node i, is the standard deviation of the output value corresponding to node i, is the deviation score of node i at time t.

[0088] In an optional implementation, the bottleneck nodes are a given number of nodes corresponding to the highest deviation scores among the multiple nodes, wherein the bottleneck nodes are used to reflect key constraints in the renewable energy consumption information.

[0089] In an optional implementation, the configured optimization model includes an objective function and constraints.

[0090] The objective function is used to indicate the minimization of the total cost, which includes the power generation cost, the wind abandonment cost and the state variable cost of the bottleneck node, and the state variable cost is calculated by the state variable and the state penalty coefficient of the bottleneck node.

[0091] In an optional implementation, the constraint condition includes at least one of the following: a power balance constraint, a thermal balance constraint, and a bottleneck node status constraint.

[0092] The bottleneck node state constraint is used to limit the state variable of the bottleneck node to be within a set safety threshold range.

[0093] In summary, the embodiments of the present application have at least the following beneficial effects: According to an embodiment of the present application, a graph structure is generated that matches the electric-thermal coupling system to be scheduled and includes multiple nodes, wherein the electric-thermal coupling system to be scheduled includes multiple components, and the components are used to indicate the renewable energy consumption information in the electric-thermal coupling system to be scheduled, and the multiple nodes correspond to the multiple components one by one; based on the data of the electric-thermal coupling system to be scheduled, the actual power value of each component is determined, and the power output value corresponding to each node is determined by using the graph attention mechanism; for each node, the deviation score of the node is determined based on the difference between the actual power value of the component corresponding to the node and the power output value corresponding to the node; a bottleneck node is determined from the multiple nodes based on the deviation score; based on the bottleneck node, an optimization model is configured to schedule the electric-thermal coupling system to be scheduled according to the configured optimization model, so as to optimize the scheduling of the electric-thermal coupling system and improve the efficiency and reliability of the scheduling of the electric-thermal coupling system.

[0094] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary hardware platform, and of course, it can also be implemented entirely by hardware. Based on such an understanding, all or part of the contribution of the technical solution of the present application to the background technology can be embodied in the form of a software product, and the computer software product can be stored in a storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present application or some parts of the embodiments.

[0095] The above is a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications are also considered to be within the scope of protection of the present application.

Claims

1. A method for dispatching an electric-thermal coupling system based on renewable energy, characterized in that: include: Generate a graph structure that matches the electric-thermal coupling system to be scheduled and includes a plurality of nodes, wherein the electric-thermal coupling system to be scheduled includes a plurality of components, the components are used to indicate renewable energy consumption information in the electric-thermal coupling system to be scheduled, and the plurality of nodes correspond to the plurality of components one by one; Based on the data of the electric-thermal coupling system to be scheduled, determining the actual power value of each component, and determining the power output value corresponding to each node by using a graph attention mechanism; For each node, determining a deviation score of the node based on a difference between an actual power value of a component corresponding to the node and a power output value corresponding to the node; determining a bottleneck node from the plurality of nodes based on the deviation score; Based on the bottleneck node, an optimization model is configured to schedule the electric-thermal coupling system to be scheduled according to the configured optimization model.

2. The method for dispatching an electric-thermal coupling system based on renewable energy according to claim 1, characterized in that: The generating of a graph structure matching the electric-thermal coupling system to be scheduled and including a plurality of nodes comprises: Model each component to generate corresponding nodes; Based on the data of each component contained in the data of the electric-thermal coupling system to be scheduled, using the node embedding technology to calculate the embedding vector of the node corresponding to each component; Based on the multiple nodes and their respective embedding vectors, an adjacency matrix of the graph structure is constructed by calculating an optimal embedding vector, thereby obtaining the graph structure.

3. The method for dispatching an electric-thermal coupling system based on renewable energy according to claim 1, characterized in that: Based on the data of the electric-thermal coupling system to be scheduled, the power output value corresponding to each node is determined by using the graph attention mechanism, including: Extracting input features of nodes corresponding to each component from data of each component included in the data of the electric-thermal coupling system to be scheduled; Determining attention weights based on input features of each of the multiple nodes; Based on the attention weight, the input features of each of the multiple nodes and the weight matrix corresponding to the graph structure, using the activation function in the graph attention mechanism, calculate and obtain the aggregate representation corresponding to each node; Based on the aggregate representation corresponding to each node, the graph attention mechanism is used to determine the power output value corresponding to each node.

4. The method for dispatching an electric-thermal coupling system based on renewable energy according to claim 3, characterized in that: The method of determining the power output value corresponding to each node using a graph attention mechanism based on the aggregate representation corresponding to each node includes: The aggregation corresponding to each node is represented as the power of each node at time t, and the power of each node at time t is input into the graph attention mechanism for iterative prediction, so that the graph attention mechanism outputs the power of each node at time t+k obtained through iterative prediction as the power output value, where t is a positive real number and k is a positive real number.

5. The method for dispatching an electric-thermal coupling system based on renewable energy according to claim 1, characterized in that: The deviation score is calculated by the following formula: in, is the actual power value of the component corresponding to node i at time t, is the power output value corresponding to node i, is the standard deviation of the output value corresponding to node i, is the deviation score of node i at time t.

6. The method for dispatching an electric-thermal coupling system based on renewable energy according to claim 1, characterized in that: The bottleneck nodes are a given number of nodes corresponding to the highest deviation scores among the multiple nodes, wherein the bottleneck nodes are used to reflect the key constraints in the renewable energy consumption information.

7. The method for dispatching an electric-thermal coupling system based on renewable energy according to claim 1, characterized in that: The configured optimization model includes an objective function and constraints; The objective function is used to indicate the minimization of the total cost, which includes the power generation cost, the wind abandonment cost and the state variable cost of the bottleneck node, and the state variable cost is calculated by the state variable and the state penalty coefficient of the bottleneck node.

8. The method for dispatching an electric-thermal coupling system based on renewable energy according to claim 7, characterized in that: The constraint conditions include at least one of the following: power balance constraint, thermal balance constraint, bottleneck node status constraint; The bottleneck node state constraint is used to limit the state variable of the bottleneck node to be within a set safety threshold range.

9. A scheduling system, characterized in that: include: A graph structure generating module, used to generate a graph structure matching the electric-thermal coupling system to be scheduled and including a plurality of nodes, wherein the electric-thermal coupling system to be scheduled includes a plurality of components, the components are used to indicate renewable energy consumption information in the electric-thermal coupling system to be scheduled, and the plurality of nodes correspond to the plurality of components one by one; A power determination module, used to determine the actual power value of each component based on the data of the electric-thermal coupling system to be scheduled, and to determine the power output value corresponding to each node by using a graph attention mechanism; A deviation scoring module is used to determine, for each node, a deviation score of the node based on a difference between an actual power value of a component corresponding to the node and a power output value corresponding to the node; A bottleneck node determination module, configured to determine a bottleneck node from the plurality of nodes based on the deviation score; The optimization scheduling module is used to configure the optimization model based on the bottleneck node, so as to schedule the electric-thermal coupling system to be scheduled according to the configured optimization model.

10. The dispatching system according to claim 9, characterized in that: Based on the data of the electric-thermal coupling system to be scheduled, the power output value corresponding to each node is determined by using the graph attention mechanism, including: Extracting input features of nodes corresponding to each component from data of each component included in the data of the electric-thermal coupling system to be scheduled; Determining attention weights based on input features of each of the multiple nodes; Based on the attention weight, the input features of each of the multiple nodes and the weight matrix corresponding to the graph structure, using the activation function in the graph attention mechanism, calculate and obtain the aggregate representation corresponding to each node; Based on the aggregate representation corresponding to each node, the graph attention mechanism is used to determine the power output value corresponding to each node.

Images