Power grid equivalent inertia probability prediction method and related device
Through the Bayesian neural network and deep neural network algorithm combined with short-term weather forecasting, a mean and variance prediction model of the grid's equivalent inertia is constructed, which solves the problem of insufficient accuracy of the grid's equivalent inertia prediction under the access of high proportion of new energy, and improves the prediction ability of grid frequency safety.
Patent Information
- Application Number
- CN202510433543.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-08
- Publication Date
- 2025-07-25
AI Technical Summary
The existing grid equivalent inertia prediction methods are insufficient in the case of high proportion of new energy access, especially in extreme weather conditions, and it is difficult to effectively evaluate the safety risks of power grid frequency.
The Bayesian neural network and deep neural network algorithm are used to combine the short-term weather forecast results to construct the mean and variance prediction model of the grid equivalent inertia. The frequency response under the grid operation mode is calculated through random sampling and time domain simulation to obtain the probability distribution function of the grid equivalent inertia.
It improves the accuracy and reliability of the prediction of equal value inertia of power grid, can provide conservative and reliable references in extreme cases, ensure the safety of the power grid frequency, and promote the absorption and utilization of new energy.
Smart Images
Figure CN120372147A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of power system analysis, and particularly relates to a method and related device for probabilistic prediction of grid equivalent inertia. Background Art
[0002] Grid equivalent inertia refers to the equivalent performance of the sum of the rotational inertias of all generators in the power system in the power grid. It is an important indicator to measure the ability of the power system to respond to load changes and maintain frequency stability. In traditional power systems, conventional energy sources such as thermal power units account for a relatively large proportion, and their rotational inertias are relatively stable. Therefore, the grid equivalent inertia is also relatively fixed. However, with the rapid development of new energy (such as wind power, photovoltaic, etc.) and the proposal of the "dual carbon" goal, the installed capacity of new energy has been continuously increasing, the structure of the power system has changed significantly, and the grid equivalent inertia has gradually shown the characteristics of dynamic change.
[0003] With the large-scale access of new energy, the reduction and increased uncertainty of the grid equivalent inertia pose new challenges to the frequency security of the power grid. To address this challenge, a series of studies and practices on grid equivalent inertia prediction have been carried out in the industry. Among them, the more advanced technical solutions mainly include: (1) Statistical analysis method based on historical data: By analyzing historical power grid operation data and new energy output data, a statistical relationship model between equivalent inertia and new energy output, load change and other factors is established to predict the future change trend of the grid equivalent inertia; (2) Machine learning prediction method: Using machine learning algorithms (such as neural networks, support vector machines, etc.) to model and predict the grid equivalent inertia. This type of method can process a large amount of complex data and automatically learn the rules in the data, improving the accuracy and reliability of the prediction; (3) Probability prediction method: Considering the uncertainty and volatility of new energy output, some studies have begun to try to introduce probability theory into grid equivalent inertia prediction, and evaluate the risk of grid frequency security by calculating the probability distribution of equivalent inertia. Although the above existing methods have made certain progress in grid equivalent inertia prediction, due to the large uncertainty and volatility of new energy output and the complexity of the power system itself, there is still room for improvement in the accuracy and calculation efficiency of the existing prediction methods. Specifically, especially under extreme weather conditions, the accuracy of the prediction results of the existing methods is greatly affected. Summary of the Invention
[0004] The object of the present invention is to provide a method and related device for predicting the probability of grid equivalent inertia, so as to solve one or more of the above existing technical problems. The technical solution disclosed by the present invention is specifically a probability prediction scheme for grid equivalent inertia considering the access of a high proportion of new energy. It predicts the probability distribution of grid equivalent inertia based on the short-term weather forecast results, so as to clarify the grid equivalent inertia boundary under different weather conditions, and can provide a reference basis for the formulation of the power system operation mode and the emergency control decision of dispatchers.
[0005] To achieve the above object, the present invention adopts the following technical solutions: In the first aspect of the present invention, a method for predicting the probability of grid equivalent inertia is provided, including the following steps: Obtain the short-term weather forecast results; Taking the obtained short-term weather forecast results as input, use the pre-trained mean and variance prediction model of grid equivalent inertia to predict, and obtain the mean and variance prediction results of grid equivalent inertia; Based on the obtained mean and variance prediction results of grid equivalent inertia, calculate to obtain the probability distribution function of grid equivalent inertia and use it as the grid equivalent inertia probability prediction result; Among them, the steps for obtaining the pre-trained mean and variance prediction model of grid equivalent inertia include: Combined with historical short-term weather forecast results, construct and train a Bayesian neural network model suitable for predicting the probability of new energy output; based on the obtained Bayesian neural network model, randomly sample to obtain a variety of feasible grid operation modes, determine the frequency response curves of the power system under different grid operation modes through time-domain simulation, and calculate the grid equivalent inertia; statistically obtain the mean and variance of the obtained grid equivalent inertia, take the historical short-term weather forecast results as input, and take the two-parameter combination of the mean and variance of grid equivalent inertia as output, and use the deep neural network algorithm to construct and train to obtain the mean and variance prediction model of grid equivalent inertia.
[0006] In a further improvement of the technical solution of the present invention, in the step of calculating the probability distribution function of grid equivalent inertia based on the obtained mean and variance prediction results of grid equivalent inertia and using it as the grid equivalent inertia probability prediction result, the calculation formula of the probability distribution function of grid equivalent inertia is: ; In the formula, is the probability density when the grid equivalent inertia takes the value of , is the possible value of the grid equivalent inertia; is the mean of the predicted grid equivalent inertia; is the variance of the predicted grid equivalent inertia; It is a mean and variance prediction model of grid equivalent inertia. It is the obtained short-term weather forecast result.
[0007] A further improvement of the technical solution of the present invention lies in that the steps of constructing and training a Bayesian neural network model suitable for new energy output probability prediction by combining historical short-term weather forecast results include: Taking the historical short-term weather forecast result as the input and the new energy output of the historical record of the new energy power station as the reference output, a Bayesian neural network model suitable for new energy output probability prediction is constructed and trained by using the Bayesian neural network algorithm.
[0008] A further improvement of the technical solution of the present invention lies in that in the step of randomly sampling to obtain a variety of feasible grid operation modes based on the obtained Bayesian neural network model, determining the frequency response curve of the power system under different grid operation modes through time-domain simulation, and calculating the grid equivalent inertia, The historical short-term weather forecast result is input into the trained Bayesian neural network model multiple times, and the new energy output of each output is recorded. And in the time-domain simulation model, power flow calculations are set and carried out multiple times. If the power flow can converge and the generator output does not exceed the limit, the voltage amplitude of each node does not exceed the limit, the line transmission power does not exceed the limit, and the power system can operate normally, then the corresponding grid operation mode is considered feasible; One thermal power unit farthest from the new energy connection point is removed, and the frequency response curve of the main network bus of the power system is obtained through transient time-domain simulation.
[0009] A further improvement of the technical solution of the present invention lies in that in the step of randomly sampling to obtain a variety of feasible grid operation modes based on the obtained Bayesian neural network model, determining the frequency response curve of the power system under different grid operation modes through time-domain simulation, and calculating the grid equivalent inertia, The calculation formula of the grid equivalent inertia is: ; In the formula, is the calculated and estimated grid equivalent inertia; is the number of main network buses of the system, are all bus number counts; is the bus weight; is the active power of the removed thermal power unit; are respectively the transient initial frequency deviation amounts obtained by simulation at the bus ; are respectively the normalized frequency deviation amounts at the bus and the bus .
[0010] A further improvement of the technical solution of the present invention lies in that, in the step of statistically obtaining the mean and variance of the grid equivalent inertia, The statistical calculation formulas for the mean and variance of the grid equivalent inertia are as follows: ; In the formula, is the mean of the grid equivalent inertia under different operation modes; is the counting number of the operation mode, is the total number of operation modes; corresponds to the operation mode the calculated value of the grid equivalent inertia; is the variance of the grid equivalent inertia under different operation modes.
[0011] A further improvement of the technical solution of the present invention lies in that the short-term weather forecast result in the step of obtaining the short-term weather forecast result has the same meteorological quantity as the historical short-term weather forecast result in the step of obtaining the prediction model of the mean and variance of the grid equivalent inertia that has been pre-trained.
[0012] In the second aspect of the present invention, a grid equivalent inertia probability prediction system is provided, including: A data acquisition module for acquiring short-term weather forecast results; A data prediction module for using the acquired short-term weather forecast result as an input, and using a pre-trained prediction model of the mean and variance of the grid equivalent inertia to perform prediction to obtain prediction results of the mean and variance of the grid equivalent inertia; A data calculation module for calculating based on the obtained prediction results of the mean and variance of the grid equivalent inertia to obtain a probability distribution function of the grid equivalent inertia and use it as a grid equivalent inertia probability prediction result; Among them, the step of obtaining the pre-trained prediction model of the mean and variance of the grid equivalent inertia includes: Combining historical short-term weather forecast results, constructing and training a Bayesian neural network model suitable for new energy output probability prediction; based on the obtained Bayesian neural network model, randomly sampling to obtain a variety of feasible grid operation modes, determining the frequency response curves of the power system under different grid operation modes through time-domain simulation, and calculating the grid equivalent inertia; statistically obtaining the mean and variance of the grid equivalent inertia, using the historical short-term weather forecast result as an input, and using the two-parameter combination of the mean and variance of the grid equivalent inertia as an output, and constructing and training a prediction model of the mean and variance of the grid equivalent inertia by using a deep neural network algorithm.
[0013] In a third aspect of the present invention, there is provided an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the program, it implements the grid equivalent inertia probability prediction method as described in any one of the first aspects of the present invention.
[0014] In a fourth aspect of the present invention, there is provided a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the grid equivalent inertia probability prediction method as described in any one of the first aspects of the present invention.
[0015] Compared with the prior art, the present invention has the following beneficial effects: For a power system with a high proportion of new energy, when the proportion of new energy output shows a continuous change, there is also a certain degree of uncertainty in the grid equivalent inertia, and the frequency security of the grid may face potential risks. In view of the problem of insufficient accuracy of the prediction results of the existing methods, the present invention specifically provides a grid equivalent inertia probability prediction method considering the access of a high proportion of new energy. By constructing a mapping from the short-term weather forecast results to the mean and variance of the grid equivalent inertia, the probability distribution function of the grid equivalent inertia under the conditions of high proportion of new energy access and meteorological condition changes can be quickly obtained, which can provide a reference and decision-making basis for the formulation of operation modes and frequency security and stability control. In summary, compared with the existing technical solutions, the technical solution of the present invention provides richer information on the grid equivalent inertia through probability prediction, can effectively ensure the conservativeness and reliability of the equivalent inertia estimation in extreme cases, can effectively improve the calculation efficiency and accuracy of the grid equivalent inertia of the power grid with a high proportion of new energy, and is of great significance for ensuring the frequency security of the grid and promoting the consumption and utilization of new energy. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art; obviously, the drawings in the following description are some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can be obtained according to these drawings without creative efforts.
[0017] Figure 1 It is a schematic flow chart of a grid equivalent inertia probability prediction method in an embodiment of the present invention; Figure 2 It is a schematic flow chart of a grid equivalent inertia probability prediction method considering the access of a high proportion of new energy in a specific embodiment of the present invention; Figure 3 It is a schematic diagram of a grid equivalent inertia probability prediction system in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0018] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention; obviously, the described embodiments of the technical solutions are a part of the embodiments of the present invention, rather than all of the embodiments.
[0019] Based on the technical solutions disclosed in 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. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.
[0020] Please refer to Figure 1 , a method for probabilistic prediction of grid equivalent inertia provided by an embodiment of the present invention includes the following steps: Step 1, obtain the short-term weather forecast result; Step 2, use the obtained short-term weather forecast result as input, and use a pre-trained mean and variance prediction model of grid equivalent inertia to perform prediction to obtain the mean and variance prediction results of grid equivalent inertia; Step 3, perform calculations based on the obtained mean and variance prediction results of grid equivalent inertia to obtain the probability distribution function of grid equivalent inertia and use it as the probabilistic prediction result of grid equivalent inertia; Among them, the steps for obtaining the pre-trained mean and variance prediction model of grid equivalent inertia include: Combine historical short-term weather forecast results to construct and train a Bayesian neural network model suitable for probabilistic prediction of new energy output; based on the obtained Bayesian neural network model, randomly sample to obtain a variety of feasible grid operation modes, determine the frequency response curves of the power system under different grid operation modes through time-domain simulation, and calculate the grid equivalent inertia; statistically obtain the mean and variance of the obtained grid equivalent inertia, use the historical short-term weather forecast result as input, and use the two-parameter combination of the mean and variance of grid equivalent inertia as output, and use a deep neural network algorithm to construct and train to obtain the mean and variance prediction model of grid equivalent inertia.
[0021] In the embodiments of the present invention, aiming at the problems existing in the prediction of grid equivalent inertia, especially the insufficient prediction accuracy caused by the uncertainty and large volatility of new energy output and the complexity of the power system, especially the influence under extreme weather conditions, a method for probabilistic prediction of grid equivalent inertia is proposed. Among them, the short-term weather forecast results are introduced, and weather factors are incorporated into the prediction model, which helps to more accurately consider the influence of weather changes on the output of new energy (such as wind power and solar energy). The Bayesian neural network can handle uncertainty. By combining historical weather forecasts and new energy output data, a model capable of predicting the probability of new energy output is trained, and this model can provide more comprehensive prediction information. By randomly sampling, a variety of feasible grid operation modes are obtained, and time-domain simulation is used to simulate the frequency response of the power system under these operation modes, so as to calculate the grid equivalent inertia. This method considers the dynamic behavior of the power system under different operating conditions and improves the prediction accuracy. Based on the predicted mean and variance, the probability distribution function of the grid equivalent inertia is calculated, providing a more intuitive prediction result, which is convenient for decision-makers to understand the uncertainty of the prediction and formulate corresponding strategies. Using the deep neural network algorithm, a model capable of directly predicting the mean and variance of the grid equivalent inertia is trained by combining historical data, improving the prediction efficiency and accuracy. In summary, the technical solution of the embodiments of the present invention effectively solves the problems existing in the existing prediction of grid equivalent inertia and improves the accuracy and reliability of the prediction results by introducing weather forecasts, using Bayesian neural networks and deep neural network algorithms, and considering the dynamic behavior of the power system.
[0022] Please refer to Figure 2 , a method for probabilistic prediction of grid equivalent inertia provided by the embodiments of the present invention includes the following steps: Step 1: Combine historical short-term weather forecast results to construct and train a Bayesian neural network model suitable for predicting the probability of new energy output; Step 2: Based on the Bayesian neural network model obtained in Step 1, randomly sample a large number of feasible grid operation modes, determine the frequency response curve of the power system under different grid operation modes through time-domain simulation, and calculate the grid equivalent inertia; Step 3: Statistically analyze the mean and variance of the grid equivalent inertia obtained in Step 2. Taking the historical short-term weather forecast results as the input and the combination of the two parameters of the mean and variance as the output, use the deep neural network algorithm to construct and train a prediction model for the mean and variance of the grid equivalent inertia; Step 4: Given the real-time short-term weather forecast results, input the prediction model for the mean and variance of the grid equivalent inertia obtained in Step 3 to infer the corresponding mean and variance, and then obtain the probability distribution function of the grid equivalent inertia.
[0023] In an embodiment of the present invention, the Bayesian neural network model in step one needs to be constructed and trained using the Bayesian neural network algorithm with the short-term weather forecast results of historical records as the input and the new energy output of the new energy power station of historical records as the reference output.
[0024] In an embodiment of the present invention, obtaining a large number of feasible power grid operation modes by random sampling in step two means inputting the short-term weather forecast results into the Bayesian neural network model obtained in step one multiple times, recording the new energy output of each output, and then setting and performing power flow calculations multiple times in the time-domain simulation model. If the power flow can converge and the generator output does not exceed the limit, the voltage amplitude of each node does not exceed the limit, the line transmission power does not exceed the limit, and the system can operate normally, then the corresponding power grid operation mode is considered feasible.
[0025] In an embodiment of the present invention, the method for obtaining the frequency response curve of the power grid in step two is to cut off a thermal power unit farthest from the new energy connection point and obtain the frequency response curve of the main network bus of the system through transient time-domain simulation.
[0026] In an embodiment of the present invention, the calculation method of the grid equivalent inertia is as follows: ; where is the calculated and estimated grid equivalent inertia, is the number of main network buses of the system, are all bus number counts, is the bus weight, is the active power of the cut-off thermal power unit, are respectively the transient initial frequency deviation amounts obtained by simulation at bus , are respectively the normalized frequency deviation amounts between bus and bus .
[0027] The statistical calculation method of the mean and variance of the grid equivalent inertia is as follows: ; where is the mean of the grid equivalent inertia under different operation modes, is the operation mode count number, is the total number of operation modes, is the calculated value of the grid equivalent inertia corresponding to operation mode , is the variance of the grid equivalent inertia under different operation modes.
[0028] The calculation formula of the probability distribution function of the grid equivalent inertia is: ; In the formula, is a possible value of the equivalent inertia of the power grid, is the probability density when the equivalent inertia of the power grid takes the value of ; is the mean value of the equivalent inertia of the power grid predicted by using the mean and variance prediction model in Step 3, is the variance of the equivalent inertia of the power grid predicted by using the mean and variance prediction model in Step 3, is the mean and variance prediction model of the equivalent inertia of the power grid obtained in Step 3, is the result of real-time short-term weather forecast.
[0029] In a specific embodiment of the present invention, the short-term weather forecast results involved in Step 1, Step 2, and Step 3 should have the same meteorological quantities as the real-time short-term weather forecast results in Step 4, including but not limited to solar zenith angle, solar azimuth angle, extraterrestrial radiation, horizontal total radiation, total irradiance on an inclined plane, wind speed, temperature, humidity, etc.
[0030] In a specific exemplary technical solution of the present invention, a method for predicting the probability of the equivalent inertia of a power grid considering high-proportion new energy access includes: Step 1: Select 12 new energy centralized access stations in a certain regional power grid in northwest China. Taking the historical short-term weather forecast results as the input and the new energy output of the new energy stations in the historical records as the reference output, use the Bayesian neural network algorithm to construct and train a Bayesian neural network model for predicting the probability of new energy output.
[0031] Step 2: Based on the Bayesian neural network model in Step 1, randomly select the short-term weather forecast results of 1000 typical meteorological scenarios, input them into the Bayesian neural network model, generate 50000 groups of feasible power grid operation modes by sampling 50 times for each typical meteorological scenario. In each mode, cut off a thermal power unit at the farthest end from the new energy connection point, obtain the frequency response curve of the main network bus of the system through transient time-domain simulation, and calculate the corresponding equivalent inertia of the power grid.
[0032] Step 3: Under each typical meteorological scenario, count the equivalent inertia of the power grid for 50 groups of power grid operation modes, and calculate the mean and variance of the equivalent inertia of the power grid. Then, taking the historical short-term weather forecast results of each typical meteorological scenario as the input and the two-parameter combination of the mean and variance of the equivalent inertia of the power grid as the output, use the deep neural network algorithm to construct and train a mean and variance prediction model of the equivalent inertia of the power grid.
[0033] Step 4: When making a prediction, given the real-time short-term weather forecast results of the meteorological scenario to be predicted, input the mean and variance prediction model of the grid equivalent inertia obtained in Step 3, and the corresponding mean and variance can be inferred, and then the probability distribution function of the grid equivalent inertia can be calculated. In a certain scenario, the predicted mean of the grid equivalent inertia is 526.74 , and the variance is 6.31. Therefore, the probability distribution function of the grid equivalent inertia in this scenario is as follows: .
[0034] In summary, a safe and reliable power supply is the foundation of social and economic development. With the proposal of the "dual carbon" goal, the installed capacity of new energy has been continuously increasing, and the equivalent inertia of the power grid has gradually decreased. When large-scale power disturbances such as sudden load increases and sudden disconnections of thermal power units occur, the frequency of the power grid may drop below the limit allowed by the safety and stability guidelines during the transient process, and the system faces the risk of frequency instability. Since the output of new energy is itself restricted by natural conditions and has significant uncertainty and volatility, for a power system with a high proportion of new energy, when the proportion of new energy output changes continuously, there is also a certain degree of uncertainty in the equivalent inertia of the power grid, and the frequency safety of the power grid may face potential risks. The method for predicting the probability of grid equivalent inertia considering the access of a high proportion of new energy provided by the embodiments of the present invention combines historical short-term weather forecast results, constructs and trains a Bayesian neural network model suitable for predicting the probability of new energy output; randomly samples a large number of feasible grid operation modes, determines the frequency response curves of the power system under different grid operation modes through time-domain simulation, and calculates the grid equivalent inertia; statistically analyzes the mean and variance of the grid equivalent inertia, constructs and trains a mean and variance prediction model of the grid equivalent inertia with historical short-term weather forecast results as the input and the mean and variance as the output; given the real-time short-term weather forecast results, the corresponding mean and variance can be inferred based on the prediction model, and then the probability distribution function of the grid equivalent inertia can be obtained. The technical solution of the embodiments of the present invention can effectively improve the calculation efficiency and accuracy of the equivalent inertia of a power grid with a high proportion of new energy, and can provide a reference for the safe and stable operation of the system and emergency decision-making after a fault.
[0035] The following is an apparatus embodiment of the present invention, which can be used to execute the method embodiment of the present invention. For details not disclosed in the apparatus embodiment, please refer to the method embodiment of the present invention.
[0036] Please refer to Figure 3 , in the embodiments of the present invention, a system for predicting the probability of grid equivalent inertia is provided, including: A data acquisition module, configured to acquire short-term weather forecast results; A data prediction module, which takes the obtained short-term weather forecast results as input, and uses a pre-trained mean and variance prediction model of the grid equivalent inertia to perform prediction, and obtains the mean and variance prediction results of the grid equivalent inertia; A data calculation module, which is used to calculate based on the obtained mean and variance prediction results of the grid equivalent inertia, obtain the probability distribution function of the grid equivalent inertia and use it as the grid equivalent inertia probability prediction result; Among them, the acquisition steps of the pre-trained mean and variance prediction model of the grid equivalent inertia include: Combined with historical short-term weather forecast results, construct and train a Bayesian neural network model suitable for new energy output probability prediction; based on the obtained Bayesian neural network model, randomly sample to obtain a variety of feasible grid operation modes, determine the frequency response curves of the power system under different grid operation modes through time-domain simulation, and calculate the grid equivalent inertia; statistically obtain the mean and variance of the grid equivalent inertia, use the historical short-term weather forecast results as input, and use the two-parameter combination of the mean and variance of the grid equivalent inertia as output, and use a deep neural network algorithm to construct and train to obtain the mean and variance prediction model of the grid equivalent inertia.
[0037] In an embodiment of the present invention, a computer device is provided. The computer device includes a processor and a memory. The memory is used to store a computer program. The computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used to execute the operations of the grid equivalent inertia probability prediction method.
[0038] In an embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a computer device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and, of course, the extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, and the operating system of the terminal is stored in this storage space. Moreover, one or more instructions suitable for being loaded and executed by the processor are stored in this storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM (Random Access Memory) or a non-volatile memory, such as at least one disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the grid equivalent inertia probability prediction method in the above embodiment.
[0039] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, optical storage, etc.) containing computer-usable program code.
[0040] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the specified functions in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.
[0041] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including an instruction device, and the instruction device implements the specified functions in the flow Figure 1One or more processes and / or boxes Figure 1 The functions specified in one or more boxes.
[0042] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 One or more processes and / or boxes Figure 1 The steps of the functions specified in one or more boxes.
[0043] Finally, it should be noted that 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 above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific implementation manners of the present invention. Any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention shall be covered by the protection scope of the claims of the present invention.
Claims
1. A method for probabilistic prediction of equivalent inertia of power grid, characterized in that, It includes the following steps: Obtain the short-term weather forecast results; Using the obtained short-term weather forecast results as input, and predicting by using a pre-trained mean and variance prediction model of the grid equivalent inertia to obtain the mean and variance prediction results of the grid equivalent inertia; Based on the obtained mean and variance prediction results of the grid equivalent inertia, calculate to obtain the probability distribution function of the grid equivalent inertia and use it as the grid equivalent inertia probability prediction result; Among them, the steps for obtaining the pre-trained mean and variance prediction model of the grid equivalent inertia include: Combined with historical short-term weather forecast results, construct and train a Bayesian neural network model suitable for new energy output probability prediction; based on the obtained Bayesian neural network model, randomly sample to obtain a variety of feasible grid operation modes, determine the frequency response curves of the power system under different grid operation modes through time-domain simulation, and calculate the grid equivalent inertia; statistically obtain the mean and variance of the grid equivalent inertia obtained, use the historical short-term weather forecast results as input, and use the two-parameter combination of the mean and variance of the grid equivalent inertia as output, and use a deep neural network algorithm to construct and train to obtain the mean and variance prediction model of the grid equivalent inertia.
2. The probability prediction method of the equivalent inertia of the power grid according to claim 1, wherein In the step of calculating the probability distribution function of the grid equivalent inertia based on the obtained mean and variance prediction results of the grid equivalent inertia and using it as the grid equivalent inertia probability prediction result, the calculation formula of the probability distribution function of the grid equivalent inertia is: ； In the formula, is the probability density when the equivalent inertia of the power grid takes the value of , is the possible value of the equivalent inertia of the power grid; is the mean value of the equivalent inertia of the power grid obtained by prediction; is the variance of the equivalent inertia of the power grid obtained by prediction; is the prediction model of the mean value and variance of the equivalent inertia of the power grid, is the short-term weather forecast result obtained.
3. A method for predicting the probability of equivalent inertia of a power grid according to claim 1, characterized in that, The steps of combining historical short-term weather forecast results, constructing and training a Bayesian neural network model suitable for new energy output probability prediction include: Using the historical short-term weather forecast results as input and the new energy output of the historical record of the new energy station as the reference output, and using the Bayesian neural network algorithm to construct and train a Bayesian neural network model suitable for new energy output probability prediction.
4. A method for predicting the probability of the equivalent inertia of a power grid according to claim 1, characterized in that In the step of randomly sampling to obtain a variety of feasible grid operation modes based on the obtained Bayesian neural network model, determining the frequency response curves of the power system under different grid operation modes through time-domain simulation, and calculating the grid equivalent inertia, Input the historical short-term weather forecast results into the trained Bayesian neural network model multiple times, record the new energy output of each output, and set and perform power flow calculations multiple times in the time-domain simulation model. If the power flow can converge and the generator output does not exceed the limit, the voltage amplitude of each node does not exceed the limit, the line transmission power does not exceed the limit, and the power system can operate normally, then the corresponding grid operation mode is considered feasible; Disconnect one thermal power unit at the farthest end from the new energy connection point, and obtain the frequency response curve of the main network bus of the power system through transient time-domain simulation.
5. A method for predicting the probability of the equivalent inertia of a power grid according to claim 4, characterized in that, In the step of randomly sampling to obtain a variety of feasible grid operation modes based on the obtained Bayesian neural network model, determining the frequency response curves of the power system under different grid operation modes through time-domain simulation, and calculating the grid equivalent inertia, The calculation formula of the grid equivalent inertia is: ； In the formula, is the calculated estimated equivalent inertia of the power grid; is the number of main network buses in the system, are all bus number counts; is the bus weight; is the active power of the thermal power unit to be cut off; are respectively the transient initial frequency deviation amounts obtained by simulation at the bus ; are respectively the normalized frequency deviation amounts at the bus and the bus .
6. A probability prediction method for the equivalent inertia of a power grid according to claim 1, characterized in that In the step of statistically obtaining the mean and variance of the grid equivalent inertia obtained, The statistical calculation formula of the mean and variance of the grid equivalent inertia is: ； In the formula, is the mean value of the equivalent inertia of the power grid under different operating modes; is the counting number of the operating mode, is the total number of operating modes; is the calculated value of the equivalent inertia of the power grid corresponding to the operating mode ; is the variance of the equivalent inertia of the power grid under different operating modes.
7. A probability prediction method for grid equivalent inertia according to claim 1, characterized in that The short-term weather forecast result in the step of obtaining the short-term weather forecast result has the same meteorological quantities as the historical short-term weather forecast result in the step of obtaining the pre-trained mean and variance prediction model of the grid equivalent inertia.
8. A power grid equivalent inertia probability prediction system, characterized in that, It includes: A data acquisition module for acquiring short-term weather forecast results; A data prediction module for using the acquired short-term weather forecast results as input and using the pre-trained mean and variance prediction model of the grid equivalent inertia to perform prediction to obtain the mean and variance prediction results of the grid equivalent inertia; A data calculation module for calculating based on the obtained mean and variance prediction results of the grid equivalent inertia to obtain the probability distribution function of the grid equivalent inertia and use it as the probability prediction result of the grid equivalent inertia; Among them, the step of obtaining the pre-trained mean and variance prediction model of the grid equivalent inertia includes: Combining historical short-term weather forecast results, constructing and training a Bayesian neural network model suitable for new energy output probability prediction; based on the obtained Bayesian neural network model, randomly sampling to obtain a variety of feasible grid operation modes, determining the frequency response curves of the power system under different grid operation modes through time-domain simulation, and calculating the grid equivalent inertia; statistically obtaining the mean and variance of the obtained grid equivalent inertia, using the historical short-term weather forecast result as input, and using the two-parameter combination of the mean and variance of the grid equivalent inertia as output, and constructing and training a mean and variance prediction model of the grid equivalent inertia using a deep neural network algorithm.
9. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the grid equivalent inertia probability prediction method according to any one of claims 1 to 7.
10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the grid equivalent inertia probability prediction method according to any one of claims 1 to 7.