Power system carbon emission flow rapid calculation method, system, equipment and medium

Abstract

The invention discloses a power system carbon emission flow rapid calculation method, system and device and a medium, and the method comprises the steps: constructing a BP neural network model for predicting a node carbon emission factor according to the calculation characteristics of a power system carbon emission factor and in combination with the complexity and nonlinear characteristics of carbon emission calculation prediction; optimizing hyper-parameters of the BP neural network model through a multi-universe optimization algorithm to obtain an MVO-BP neural network model; inputting the preprocessed real-time dynamic operation parameters of the power system into the MVO-BP neural network model, and outputting carbon emission factors of nodes; and according to the carbon emission factor of the node, calculating to obtain a branch carbon emission flow. Therefore, the problem that carbon emission factor calculation is complex due to the fact that the prior art depends on power grid operation parameters and power generation load conditions is solved; and the method has a post-evaluation characteristic and cannot reflect the dynamic change of the power system in real time.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon emission monitoring, and in particular, to a method, system, device and medium for rapidly calculating carbon emission flow in a power system. Background Art

[0002] Currently, global climate change has become a major challenge faced by countries around the world. To achieve the sustainable development goals, many countries have successively formulated greenhouse gas emission reduction policies. As one of the main sources of carbon emissions, the power industry accounts for about 30%-40% of the global total emissions and plays a key role in realizing the energy structure transformation and carbon emission reduction goals. And the calculation ability of dynamic carbon emission factors has become an important indicator for measuring the low-carbon development level of power systems.

[0003] At present, the power carbon emission calculation method based on power flow has become one of the common technical means in power system carbon emission analysis. However, the carbon emission calculation method based on power flow has some limitations. First, the power flow calculation itself has a large amount of calculation when facing complex power grids or large-scale systems, and is greatly affected by the system state, with poor real-time performance; second, since the power flow calculation depends on the dispatching strategy of generating units and the change of system load, without the support of an efficient dispatching optimization strategy, it may be difficult to fully reflect the volatility and time-variation of carbon emission flow in the actual power system. In addition, traditional power flow calculation methods often adopt relatively simplified models and may not be able to effectively capture more complex non-linear relationships in system operation. Summary of the Invention

[0004] The present invention provides a method, system, device and medium for rapidly calculating carbon emission flow in a power system, which is used to solve the problems that the existing technology depends on grid operation parameters and power generation load conditions, resulting in complex calculation of carbon emission factors; and has a post-evaluation characteristic and cannot reflect the dynamic changes of the power system in real time.

[0005] In view of this, the first aspect of this application provides a method, which includes:

[0006] Construct a BP neural network model for predicting node carbon emission factors according to the calculation characteristics of power system carbon emission factors and combining the complexity and non-linear characteristics of carbon emission calculation prediction;

[0007] Optimize the hyperparameters of the BP neural network model through the multi-universe optimization algorithm to obtain an MVO-BP neural network model;

[0008] Input the preprocessed real-time dynamic operation parameters of the power system into the MVO-BP neural network model and output the carbon emission factors of the nodes;

[0009] The branch carbon emission flow is calculated based on the carbon emission factors of the nodes.

[0010] Optionally, according to the calculation characteristics of the carbon emission factors of the power system and in combination with the complexity and nonlinear characteristics of carbon emission calculation prediction, a BP neural network model for predicting the carbon emission factors of nodes is constructed, including:

[0011] According to the calculation characteristics of the carbon emission factors of the power system, key features affecting the carbon emission factors of nodes are selected as input variables, and the carbon emission factors of the nodes are used as output variables;

[0012] The number of neurons in the input layer is designed to be the same as the number of the key features, and each input node corresponds to one of the key features;

[0013] According to the complexity and nonlinear characteristics of carbon emission calculation prediction, a network structure including multiple hidden layers is designed, and the ReLU activation function is adopted, thereby constructing a BP neural network model for predicting the carbon emission factors of nodes.

[0014] Optionally, the hyperparameters of the BP neural network model are optimized by the multi-universe optimization algorithm to obtain an MVO-BP neural network model, including:

[0015] A group of initial parameters of the BP neural network are represented by universes, the loss function of the training process of the BP neural network is defined according to the universes, the parameters of the BP neural network model are optimized, and the optimized optimal parameters are used to initialize the BP neural network and train it, thereby obtaining an MVO-BP neural network model.

[0016] Optionally, the calculation of the branch carbon emission flow according to the carbon emission factors of the nodes includes:

[0017] Substitute the carbon emission factors of the nodes into the branch carbon emission flow calculation formula to calculate the branch carbon emission flow;

[0018] Among them, the branch carbon emission flow calculation formula is:

[0019] ;

[0020] In the formula, is the carbon emission factor of node , is the power flow of the branch from node i to node j.

[0021] The second aspect of this application provides a power system carbon emission flow rapid calculation system, and the system includes:

[0022] A construction unit, which is used to construct a BP neural network model for predicting the carbon emission factor of nodes according to the calculation characteristics of the carbon emission factor of the power system and in combination with the complexity and nonlinear characteristics of carbon emission calculation and prediction;

[0023] An optimization unit, which is used to optimize the hyperparameters of the BP neural network model through a multi-universe optimization algorithm to obtain an MVO-BP neural network model;

[0024] A prediction unit, which is used to input the real-time dynamic operation parameters of the power system after preprocessing into the MVO-BP neural network model and output the carbon emission factor of the node;

[0025] A calculation unit, which is used to calculate the branch carbon emission flow according to the carbon emission factor of the node.

[0026] Optionally, the construction unit is specifically used for:

[0027] According to the calculation characteristics of the carbon emission factor of the power system, select the key features that affect the carbon emission factor of the node and use them as input variables, and use the carbon emission factor of the node as the output variable;

[0028] Design the number of neurons in the input layer to be the same as the number of the key features, and each input node corresponds to one of the key features;

[0029] According to the complexity and nonlinear characteristics of carbon emission calculation and prediction, design a network structure including multiple hidden layers and adopt the ReLU activation function, so as to construct a BP neural network model for predicting the carbon emission factor of nodes.

[0030] Optionally, the optimization unit is specifically used for:

[0031] Represent a set of initial parameters of the BP neural network through the universe, define the loss function of the training process of the BP neural network according to the universe, optimize the parameters of the BP neural network model, initialize the BP neural network with the optimized optimal parameters and conduct training, so as to obtain an MVO-BP neural network model.

[0032] Optionally, the calculation unit is specifically used for:

[0033] Substitute the carbon emission factor of the node into the branch carbon emission flow calculation formula to calculate the branch carbon emission flow;

[0034] Among them, the branch carbon emission flow calculation formula is:

[0035] ;

[0036] In the formula, is the carbon emission factor of node , is the power flow of the branch from node i to node j.

[0037] A third aspect of the present invention provides a device for quickly calculating the carbon emission flow of a power system, the device including a processor and a memory:

[0038] The memory is used for storing program codes and transmitting the program codes to the processor;

[0039] The processor is used for executing the steps of the method for quickly calculating the carbon emission flow of the power system as described in the first aspect above according to the instructions in the program codes.

[0040] A fourth aspect of the present invention provides a computer-readable storage medium, which is used for storing program codes, and the program codes are used for executing the method for quickly calculating the carbon emission flow of the power system as described in the first aspect above.

[0041] It can be seen from the above technical solutions that the present invention has the following advantages:

[0042] 1. Based on learning the non-linear mapping relationship between key features and carbon emission factors from historical data, the present invention realizes the rapid prediction of carbon emission flow, avoids the strong dependence of traditional methods on accurate power grid parameters and load data, and reduces the calculation complexity.

[0043] 2. Compared with the post-evaluation characteristics of traditional methods, by receiving dynamic input data of the power system in real time and quickly outputting carbon emission factors and carbon emission flow, the present invention can timely reflect the system operation state and support real-time carbon emission management and low-carbon scheduling decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0045] Figure 1 is a schematic flowchart of a method for quickly calculating the carbon emission flow of a power system provided by an embodiment of the present invention;

[0046] Figure 2 is a schematic structural diagram of a system for quickly calculating the carbon emission flow of a power system provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] To make the objectives, features, and advantages of the present invention more obvious and understandable, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the embodiments described below are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0048] Please refer to Figure 1 , a fast calculation method for carbon emission flow of a power system provided in an embodiment of the present invention includes:

[0049] Step 101: According to the calculation characteristics of the carbon emission factor of the power system and in combination with the complexity and nonlinear characteristics of carbon emission calculation and prediction, construct a BP neural network model for predicting the carbon emission factor of nodes.

[0050] Step 102: Optimize the hyperparameters of the BP neural network model through the multi-universe optimization algorithm to obtain the MVO-BP neural network model.

[0051] Step 103: Input the preprocessed real-time dynamic operation parameters of the power system into the MVO-BP neural network model and output the carbon emission factor of the nodes.

[0052] Step 104: Calculate the carbon emission flow of the branch according to the carbon emission factor of the node.

[0053] In one embodiment, Step 101 includes the following steps:

[0054] Step 1011: According to the calculation characteristics of the carbon emission factor of the power system, select the key features affecting the carbon emission factor of the node as the input variables, and use the carbon emission factor of the node as the output variable.

[0055] It should be noted that the calculation of the carbon emission factor requires the power flow of the power grid, the power generation power, and the type of power plant, etc. The calculation method is as follows:

[0056] (1)

[0057] Among them, represents the power carbon emission coefficient of the th node; represents the active power output of the power plant connected to the th consumer node; is the carbon emission coefficient of the power plant; is the active power input from the th node to the th node; is the A set of branch connection nodes of a node; is the carbon emission factor of the node.

[0058] According to the calculation characteristics of the carbon emission factor of the power system, select the key features that affect the carbon emission factor of the node as input variables: generator carbon emission factor and node load power . Output variable: carbon emission factor of each node .

[0059] Step 1012: Design the number of neurons in the input layer to be the same as the number of the key features, and each input node corresponds to one of the key features.

[0060] It should be noted that the number of neurons in the input layer is designed to be the same as the number of input features, and each input node corresponds to one feature.

[0061] Step 1013: According to the complexity and non-linear characteristics of carbon emission calculation prediction, design a network structure including multiple hidden layers, and adopt the ReLU activation function, so as to construct a BP neural network model for predicting the carbon emission factor of the node.

[0062] It should be noted that according to the complexity and non-linear characteristics of carbon emission calculation prediction, design a network structure including multiple hidden layers, and adopt the Rectified Linear Unit (ReLU) activation function to enhance the model's fitting ability for non-linear relationships. The output layer outputs multiple nodes to predict the carbon emission factor of the node .

[0063] In one embodiment, step 102 includes the following steps:

[0064] Represent a set of initial parameters of the BP neural network through the universe, define the loss function of the training process of the BP neural network according to the universe, optimize the parameters of the BP neural network model, initialize the BP neural network with the optimized optimal parameters and perform training, so as to obtain the MVO-BP neural network model.

[0065] It should be noted that in order to improve the generalization ability and accuracy of the BP neural network, the multi-universe optimization algorithm (MVO) is used to optimize the model hyperparameters. MVO constructs a mathematical model of the MVO algorithm by introducing function formulas through the motion behavior of the multi-universe population. In the constructed mathematical model, a universe is regarded as a solution to the optimization problem, and each object in a single universe is regarded as a component of the corresponding solution. The expansion rate of a single universe is proportional to the objective function value of the corresponding solution. The initialization of the solution is used to start the optimization process, and the update of each group of solutions is performed through progressive iteration rules.

[0066] Update the parameter positions of each universe according to the gravitational mechanism of the MVO algorithm:

[0067] (2)

[0068] where is the current universe position, is the current best universe position, is a random number, is the weighting factor that controls the gravitational strength.

[0069] Determine the current optimal universe as the black hole, which attracts other universes to approach it; if some universes exceed the search range, randomly re-initialize their positions. Terminate the optimization process according to the convergence of the fitness function or when the preset number of iterations is reached.

[0070] The following is an introduction to the multi-verse optimization algorithm and the process of using the MVO algorithm to optimize the calculation of carbon emission flow in the power system:

[0071] The multi-verse optimization algorithm (MVO) is inspired by the multi-verse theory in cosmology. The algorithm regards candidate solutions as different universes and evolves the solutions by simulating the process of energy (or inflation rate) migration and balance among multi-verses. MVO mainly relies on the analogy mechanisms of "white hole", "black hole" and "wormhole" to achieve global search and local exploitation in the solution space: the white hole represents the high-energy region where better solutions are located, which can release information (variable values) outward, so as to spread excellent characteristics to other universes. The black hole represents the low-energy region where worse solutions are located, which can absorb information from the white hole, thus continuously improving the quality of its own solutions. The wormhole provides a high-dimensional transition channel for solutions in the solution space, enabling solutions to make large-span jumps between different dimensions and positions, thus avoiding falling into local optima.

[0072] The MVO algorithm gradually transitions from global large-scale exploration (diversity search) to refined exploitation (local search) of the optimal solution region by dynamically changing the weights and scopes of action of each mechanism during iteration, thereby achieving fast and stable optimization performance.

[0073] 1.1 Exploration stage:

[0074] At the initial stage of the algorithm, MVO obtains sufficient information in the solution space through large-scale search and extensive migration to avoid falling into local optima prematurely. This stage mainly relies on the white hole and wormhole mechanisms. The dimension of the decision variables of the problem is d, and the population contains n universes. Each universe is represented as:

[0075] (5)

[0076] For each universe , calculate its fitness or objective function value These values ​​are sorted and normalized to give each universe's "Inflation Rate" (IR) or Normalized Inflation Rate (NI).

[0077] Excellent solutions (high IR values) are equivalent to white holes, which can spread their characteristics to other universes. When the j-th dimension variable is , a universe with high IR is selected using the following probability mechanism For reference:

[0078] (6)

[0079] in, is the normalized inflation rate of the rth universe, is a random number. This process ensures that the information of excellent solutions has a greater probability of being spread, thus improving the quality of the global solution.

[0080] In order to avoid falling into the local optimum, MVO introduces wormholes in the exploration phase to achieve leapfrogging of solutions. The wormhole existence probability WEP (Wormhole Existence Probability) and the space-time distortion rate TDR (Travel Distance Rate) are defined and usually adjusted dynamically with iterations:

[0081] (7)

[0082] (8)

[0083] When a solution migrates through a wormhole, its position refers to the global optimal solution. Make a random perturbation:

[0084] (9)

[0085] in, and are the upper and lower boundaries of the j-th dimension variable, is a random number in [-1,1]. When WEP is high, wormhole jumps frequently, which is conducive to global exploration. Through the combined effect of white hole and wormhole mechanism, the exploration phase can ensure that sufficiently diverse and dispersed solutions are obtained in the solution space, providing parameters for constructing BP neural network to quickly calculate carbon emission flow of power system.

[0086] 1.2 Transformation from exploration to development:

[0087] As the iteration progresses, the algorithm gradually accumulates a global understanding of the solution space. In the middle and late stages, MVO needs to shift from large-scale exploration to local refinement of the relatively optimal solution region. At this time, the algorithm dynamically adjusts the parameters of each mechanism, making the changing trends of parameters such as WEP converge, reducing large-scale random jumps, and instead strengthening the optimization depth within the local domain. This transformation is mainly reflected in the following aspects:

[0088] (1) Parameter regulation: As the iteration increases, WEP increases from a lower value to a higher value, but tends to stabilize when approaching the maximum number of iterations; TDR gradually decreases, indicating that more attention is paid to fine-tuning in a small range in the later stage.

[0089] (2) Change in search tendency: In this stage, the algorithm gradually reduces the exploration of regions far from the global optimal solution, and invests more computational resources near the current relatively optimal solution to quickly approach the global optimal or approximate global optimal solution.

[0090] Through these progressive adjustments of mechanisms and parameters, MVO achieves a smooth transition from broad-spectrum search (Exploration) to fine optimization (Exploitation).

[0091] 1.3 Development stage:

[0092] In the final iteration stage, MVO mainly conducts fine development of excellent solutions through the black hole mechanism. Black holes correspond to those high-energy universes close to the optimal solution. By continuously absorbing surrounding information and narrowing the search range, the solution is further converged near the optimal position.

[0093] When the algorithm enters the development stage, the black hole mechanism begins to dominate the search process. At this time, the diffusion from white holes and the long-distance jumps through wormholes are significantly reduced, and the search is concentrated on the current optimal solution and its neighbors. The black hole mechanism can achieve local optimization by reducing the deviation between the solution and the optimal solution:

[0094] (10)

[0095] Among them, is a perturbation within a small range, used for fine-tuning around the optimal solution. In this way, the solution approaches the optimal value with higher precision.

[0096] As the development stage continues, the diversity of solutions gradually decreases, and the algorithm continuously converges to the optimal solution region. Once the objective function value or its change rate meets the preset convergence conditions (such as the maximum number of iterations, the fitness change is less than a certain threshold, etc.), the algorithm terminates.

[0097] By meticulously refining the solution in the later stage, MVO can effectively improve the accuracy and stability of the final solution, ensuring the excellence and repeatability of the optimization results.

[0098] 2. The MVO algorithm optimizes the calculation process of carbon emission flow in the power system:

[0099] Step 1: Load data, extract input and output features, and divide the data into a training set and a test set. Initialize the population size N and the maximum number of iterations T;

[0100] Step 2: Initialize the population using Equation (5), randomly generate N universes, and each universe represents a set of initial parameters of the BP neural network;

[0101] Step 3: Calculate the fitness value of each universe using the training data, defined as the mean square error MSE;

[0102] Step 4: Update the position of each universe according to the gravitational mechanism Equation (2) of the MVO algorithm;

[0103] Step 5: Introduce a wormhole using Equations (6 - 9) to achieve a leapfrog jump of the solution and avoid falling into a local optimum;

[0104] Step 6: Judge convergence using Equation (10). If it converges, stop the iteration; otherwise, return to Step 3;

[0105] Step 7: Train the optimized BP neural network. Initialize the BP neural network with the optimal parameters obtained by MVO and train it. The loss function during the training process is MSE;

[0106] Step 8: Quickly calculate the carbon emission flow according to Equations (3 - 4). Input the real - time input data into the optimized BP neural network model to quickly predict the node carbon emission factor. Calculate the branch carbon emission flow according to the carbon emission factor.

[0107] In one embodiment, Step 104 includes the following steps:

[0108] Substitute the carbon emission factor of the node into the branch carbon emission flow calculation formula to calculate the branch carbon emission flow;

[0109] Among them, the branch carbon emission flow calculation formula is:

[0110] ;

[0111] In the formula, is the carbon emission factor of node , is the power flow of the branch from node i to node j.

[0112] It should be noted that the output power of the generator is obtained in real - time through the power system data acquisition device and the load power Dynamic operation data such as... Input the preprocessed real-time data into the optimized MVO-BP neural network model to quickly calculate the carbon emission factors of each node. The branch carbon flow is based on the carbon emission factor and the power flow distribution, representing the carbon emissions flowing on the branch. The power flow from node i to node j of the branch is The branch carbon flow is calculated as follows:

[0113] (3)

[0114] (4)

[0115] Where is the carbon emission factor of node .

[0116] A fast calculation method for carbon emission flow in a power system provided in an embodiment of the present invention combines the global optimization ability of the multi-verse optimization algorithm (MVO) with the non-linear fitting advantage of the BP neural network. Without fully relying on grid operation parameters, the present invention quickly establishes a prediction model for carbon emission factors through learning and optimization of key features, realizing accurate prediction and efficient calculation of future carbon emission factors of users. Further, by receiving dynamic input data of the power system in real time and quickly outputting carbon emission factors and carbon emission flows, it can timely reflect the system operation state and support real-time carbon emission management and low-carbon scheduling decisions. Compared with the traditional genetic algorithm (GA) and particle swarm optimization algorithm (PSO), the MVO algorithm has the advantages of simple framework, few control parameters, strong adaptability, and high search efficiency. Thus, it solves the problems that the existing technology depends on grid operation parameters and power generation load conditions, resulting in complex calculation of carbon emission factors, and has the characteristics of post-evaluation and cannot reflect the dynamic changes of the power system in real time.

[0117] The above is a fast calculation method for carbon emission flow in a power system provided in an embodiment of the present invention. The following is a fast calculation system for carbon emission flow in a power system provided in an embodiment of the present invention.

[0118] Please refer to Figure 2 , a fast calculation system for carbon emission flow in a power system provided in an embodiment of the present invention includes:

[0119] A construction unit 201, configured to construct a BP neural network model for predicting the carbon emission factor of a node according to the calculation characteristics of the carbon emission factor of the power system and in combination with the complexity and non-linear characteristics of carbon emission calculation and prediction;

[0120] An optimization unit 202, configured to optimize the hyperparameters of the BP neural network model through the multi-verse optimization algorithm to obtain an MVO-BP neural network model;

[0121] A prediction unit 203, configured to input the real-time dynamic operation parameters of the preprocessed power system into the MVO-BP neural network model and output the carbon emission factor of the node;

[0122] A calculation unit 204, configured to calculate the branch carbon emission flow according to the carbon emission factor of the node.

[0123] Furthermore, in an embodiment of the present invention, there is also provided a device for quickly calculating the carbon emission flow of a power system. The device includes a processor and a memory:

[0124] The memory is used to store program codes and transmit the program codes to the processor;

[0125] The processor is configured to execute the steps of the method for quickly calculating the carbon emission flow of the power system as described in the foregoing method embodiment according to the instructions in the program codes.

[0126] Furthermore, in an embodiment of the present invention, there is also provided a computer-readable storage medium. The computer-readable storage medium is used to store program codes, and the program codes are used to execute the method for quickly calculating the carbon emission flow of the power system as described in the foregoing method embodiment.

[0127] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0128] In several embodiments provided by the present invention, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the couplings or direct couplings or communication connections shown or discussed with each other can be through some interfaces, and the indirect couplings or communication connections of devices or units can be in electrical, mechanical, or other forms.

[0129] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0130] In addition, in each embodiment of the present invention, each functional unit may be integrated into a processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of a software functional unit.

[0131] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, may be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in each embodiment of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0132] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A method for quickly calculating carbon emission flow of a power system, characterized in that: include: According to the calculation characteristics of the carbon emission factor of the power system and combined with the complexity and nonlinear characteristics of carbon emission calculation and prediction, a BP neural network model for predicting node carbon emission factors is constructed; The hyperparameters of the BP neural network model are optimized by a multi-universe optimization algorithm to obtain an MVO-BP neural network model; Input the preprocessed real-time dynamic operation parameters of the power system into the MVO-BP neural network model, and output the carbon emission factor of the node; The branch carbon emission flow is calculated according to the carbon emission factor of the node.

2. The method for rapid calculation of carbon emission flow of power system according to claim 1, characterized in that: According to the calculation characteristics of the carbon emission factor of the power system and combined with the complexity and nonlinear characteristics of carbon emission calculation prediction, a BP neural network model for predicting the node carbon emission factor is constructed, including: According to the calculation characteristics of the carbon emission factor of the power system, the key features that affect the carbon emission factor of the node are selected as input variables, and the carbon emission factor of the node is used as the output variable; Design the number of neurons in the input layer to be consistent with the number of the key features, and each input node corresponds to one of the key features; According to the complexity and nonlinear characteristics of carbon emission calculation and prediction, a network structure with multiple hidden layers is designed, and the ReLU activation function is adopted to construct a BP neural network model for predicting node carbon emission factors.

3. The method for rapid calculation of carbon emission flow of power system according to claim 1, characterized in that: The hyperparameters of the BP neural network model are optimized by the multi-universe optimization algorithm to obtain the MVO-BP neural network model, including: A set of initial parameters of the BP neural network is represented by the universe, and the loss function of the training process of the BP neural network is defined according to the universe. The parameters of the BP neural network model are optimized, and the BP neural network is initialized with the optimized optimal parameters and trained to obtain the MVO-BP neural network model.

4. The method for rapid calculation of carbon emission flow of power system according to claim 1, characterized in that: The calculating the branch carbon emission flow according to the carbon emission factor of the node includes: Substituting the carbon emission factor of the node into the branch carbon emission flow calculation formula to calculate the branch carbon emission flow; The branch carbon emission flow calculation formula is: ； In the formula, For Node The carbon emission factor, is the power flow from node i to node j in the branch.

5. A rapid calculation system for carbon emission flow in a power system, characterized in that: include: A construction unit is used to construct a BP neural network model for predicting node carbon emission factors according to the calculation characteristics of the carbon emission factor of the power system and in combination with the complexity and nonlinear characteristics of carbon emission calculation prediction; An optimization unit, used for optimizing the hyperparameters of the BP neural network model by a multi-universe optimization algorithm to obtain an MVO-BP neural network model; A prediction unit, used to input the pre-processed real-time dynamic operation parameters of the power system into the MVO-BP neural network model, and output the carbon emission factor of the node; A calculation unit is used to calculate the branch carbon emission flow according to the carbon emission factor of the node.

6. The power system carbon emission flow rapid calculation system according to claim 5 is characterized in that: The building block is specifically used for: According to the calculation characteristics of the carbon emission factor of the power system, the key features that affect the carbon emission factor of the node are selected as input variables, and the carbon emission factor of the node is used as the output variable; Design the number of neurons in the input layer to be consistent with the number of the key features, and each input node corresponds to one of the key features; According to the complexity and nonlinear characteristics of carbon emission calculation and prediction, a network structure with multiple hidden layers is designed, and the ReLU activation function is adopted to construct a BP neural network model for predicting node carbon emission factors.

7. The power system carbon emission flow rapid calculation system according to claim 5, characterized in that: The optimization unit is specifically used for: A set of initial parameters of the BP neural network is represented by the universe, and the loss function of the training process of the BP neural network is defined according to the universe. The parameters of the BP neural network model are optimized, and the BP neural network is initialized with the optimized optimal parameters and trained to obtain the MVO-BP neural network model.

8. The power system carbon emission flow rapid calculation system according to claim 5, characterized in that: The computing unit is specifically used for: Substituting the carbon emission factor of the node into the branch carbon emission flow calculation formula to calculate the branch carbon emission flow; The branch carbon emission flow calculation formula is: ； In the formula, For Node The carbon emission factor, is the power flow from node i to node j in the branch.

9. A device for rapid calculation of carbon emission flow in power system, characterized in that: The device comprises a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the method for quickly calculating carbon emission flow of a power system according to any one of claims 1 to 4 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store program codes, and the program codes are used to execute the method for quickly calculating carbon emission flow of a power system according to any one of claims 1-4.

Images