Method and system for controlling steady-state operation of power system driven by uncertain data

Abstract

The invention discloses a power system steady-state operation control method and system driven by uncertain data, and the method comprises the steps: collecting historical operation data of a new energy power system, and carrying out the data preprocessing of the historical operation data, and obtaining first power data; inputting the first power data into a preset GAN enhancement model, obtaining corresponding analog data based on a generator, and inputting the analog data and the first power data as input data into a discriminator to obtain second power data similar to the first power data, the GAN enhancement model comprises the generator and the discriminator; and inputting the second power data into a preset dynamic optimization strategy network to obtain an optimal control strategy of the power system, and controlling the power system based on the optimal control strategy. According to the invention, the steady-state operation condition of the power system can be accurately and efficiently evaluated, and the steady-state operation of the power system after the new energy is accessed is ensured.

Description

Technical Field

[0001] The present application relates to the field of steady-state control, and in particular to a method and system for controlling steady-state operation of a power system driven by uncertainty data. Background Art

[0002] With the rapid progress of new energy generation technologies such as wind power and solar energy and their large-scale integration into the power grid, traditional power systems are facing unprecedented new challenges. This type of new energy generation method is highly dependent on natural conditions, and therefore exhibits greater volatility and discontinuity, making it difficult to accurately estimate the power output, resulting in a significant increase in the dispatch complexity of the power system, and also posing a severe challenge to the stable operation status and overall safety of the system.

[0003] The traditional steady-state analysis method of the power system is based on a deterministic model, that is, it evaluates the steady-state operation of the system by assuming that all parameters in the system are known and unchanged. However, in the environment of renewable energy grid connection, the operating parameters of the power system are often in dynamic change, which brings significant uncertainty to the operation of the system. This uncertainty makes it difficult for traditional steady-state assessment methods to accurately predict the state and operating risks of the system, and it is difficult to cope with fluctuations within the system, which leads to a series of problems such as grid frequency fluctuations, voltage over-limit and even system collapse.

[0004] Application Contents

[0005] The present application provides an uncertainty data-driven steady-state operation control method and system for a power system, so as to accurately and efficiently evaluate the steady-state operation of the power system and ensure the steady-state operation of the power system after the access of new energy.

[0006] In a first aspect, the present application provides an uncertainty data-driven steady-state operation control method for a power system, comprising:

[0007] Collecting historical operation data of the new energy power system and performing data preprocessing on the historical operation data to obtain first power data, wherein the historical operation data includes voltage, frequency and power;

[0008] Inputting the first power data into a preset GAN enhanced model, obtaining corresponding simulation data based on a generator, and inputting the simulation data and the first power data into a discriminator as input data to obtain second power data similar to the first power data, wherein the GAN enhanced model includes the generator and the discriminator;

[0009] The second power data is input into a preset dynamic optimization strategy network to obtain an optimal control strategy for the power system, and the power system is controlled based on the optimal control strategy.

[0010] The embodiment of the present application can remove noise or outliers that may be contained in the historical operation data by performing data preprocessing on the collected historical operation data, so as to facilitate the subsequent GAN enhanced model to accurately obtain the second power data that is similar to the first power data and of higher quality; after learning the first power data through the generator in the GAN enhanced model, simulated data is generated, and the simulated data and the first power data are distinguished through the discriminator, so as to facilitate the second power data that is similar to the first power data and of higher quality, which can reduce the impact of uncertainty data on subsequent stability analysis, and can provide more reliable input data for subsequent steady-state operation analysis and optimal control strategy, thereby ensuring the stable operation of the power system after the new energy is connected; through the preset dynamic optimization strategy network, the steady-state operation of the power system can be accurately and efficiently evaluated based on the second power data, and the control strategy that is most suitable for the second power data can be generated, thereby ensuring the steady-state operation of the power system after the new energy is connected.

[0011] Furthermore, the historical operation data is preprocessed to obtain the first power data, specifically:

[0012] Performing wavelet transform on the historical operation data to obtain sub-signals of different frequencies, wherein the sub-signals include low-frequency signals and high-frequency signals;

[0013] Determining whether the high-frequency signal is higher than a preset threshold;

[0014] If the high-frequency signal is higher than a preset threshold, the first high-frequency signal higher than the preset threshold is removed; if the high-frequency signal is lower than the preset threshold, the second high-frequency signal lower than the preset threshold is assigned a value of zero or soft threshold processing to obtain a third high-frequency signal;

[0015] The third high-frequency signal and the low-frequency signal are reconstructed to obtain the first power data.

[0016] In this way, by preprocessing the collected historical operation data through wavelet transform, noise or outliers that may be contained in the historical operation data can be removed, making it easier for the subsequent GAN enhancement model to accurately obtain second power data that is similar to the first power data and of higher quality.

[0017] Furthermore, the historical operation data is preprocessed to obtain the first power data, specifically:

[0018] Performing empirical mode decomposition on the historical operation data to obtain a first IMF component and a residual signal;

[0019] Determine whether the first IMF component meets a preset condition, and if so, remove the second IMF component that meets the preset condition to obtain a plurality of third IMF components;

[0020] The remaining several third IMF components are added together to obtain the first power data.

[0021] In this way, by performing data preprocessing on the collected historical operation data through empirical mode decomposition, noise or outliers that may be contained in the historical operation data can be removed, making it easier for the subsequent GAN enhanced model to accurately obtain second power data that is similar to the first power data and of higher quality.

[0022] Furthermore, the first power data is input into a preset GAN enhancement model, and corresponding simulation data is obtained based on the generator, specifically:

[0023] performing standardization processing on the first power data to obtain third power data;

[0024] Encoding the third power data by a variational autoencoder to obtain corresponding latent variables;

[0025] The generator obtains simulation data corresponding to the first power data based on the latent variables and random noise.

[0026] In this way, the potential distribution characteristics of the first power data are captured by the variational autoencoder, thereby enhancing the initial representation ability of the generator to obtain latent variables, so that the generator can generate simulated data similar to the real data distribution based on these variables, thereby improving the quality and accuracy of the generated data.

[0027] Furthermore, the analog data and the first power data are input into a discriminator as input data to obtain second power data similar to the first power data, specifically:

[0028] Scaling or segmenting the simulation data and the first power data at different scales to obtain corresponding eigenvalues;

[0029] The discriminator discriminates the feature value at each scale to obtain a discriminant probability at the scale;

[0030] The discrimination probabilities at all scales are comprehensively evaluated to obtain a final discrimination probability, and second power data similar to the first power data is determined based on the final discrimination probability.

[0031] In this way, the discriminator performs multiple discriminations through feature representations at different scales, ensuring that the generator can generate data distributions close to the real one at different time scales, and then obtain second power data similar to the first power data, which can provide more reliable input data for subsequent steady-state operation analysis and optimal control strategies, thereby ensuring the stable operation of the power system after the access of new energy.

[0032] Furthermore, the objective function of the discriminator is specifically:

[0033]

[0034] Where, X real is the real first power data, D(X real ) is the discrimination probability of the first power data, X fake is the simulation data, D(X fake ) is the discrimination probability of the simulated data, p data (X) is the actual distribution of the first power data, p G (X) is the distribution of the simulation data.

[0035] Furthermore, the second power data is input into a preset dynamic optimization strategy network to obtain an optimal control strategy for the power system, specifically:

[0036] determining a first control strategy for the power system based on the second power data;

[0037] The stability evaluation result is obtained by analyzing the stability operation state of the new energy power system after the first control strategy is implemented through the dynamic optimization strategy network;

[0038] Based on the stability evaluation result, the control strategy is adjusted through the dynamic optimization strategy network to determine the optimal control strategy for the power system.

[0039] Through the preset dynamic optimization strategy network, the steady-state operation of the power system can be accurately and efficiently evaluated based on the second power data, and the control strategy that best suits the second power data can be generated, thereby ensuring the steady-state operation of the power system after the access of new energy.

[0040] Furthermore, the objective function of the dynamic optimization strategy network is:

[0041]

[0042] Where θ is the control strategy; V i and F i are the voltage and frequency of the power system at time t; V ref is the voltage reference value of the power system; F ref is the frequency reference value of the power system; P imbalance (t) is the power imbalance of the power system at time t; α is a control parameter, which represents the weight of the power imbalance.

[0043] Furthermore, after the power system is controlled based on the optimal control strategy, the method further includes:

[0044] The operating data of the new energy power system acquired in real time is analyzed through the edge computing node to obtain the analysis results;

[0045] The edge computing node dynamically adjusts the control strategy based on the analysis result, and sends the adjusted second control strategy to the power grid dispatching center, so that the power grid dispatching center determines and executes the third control strategy.

[0046] In this way, the real-time operation data of the power system is received and analyzed by the edge computing nodes in real time, and fed back to the grid dispatching center to ensure that the power system can respond quickly when the renewable energy fluctuates violently, thereby maintaining the steady-state operation of the renewable energy power system.

[0047] In a second aspect, the present application provides an uncertainty data-driven steady-state operation control system for a power system, comprising: an acquisition module, a processing module, and a control module;

[0048] The acquisition module is used to collect historical operation data of the new energy power system and perform data preprocessing on the historical operation data to obtain first power data, wherein the historical operation data includes voltage, frequency and power;

[0049] The processing module is used to input the first power data into a preset GAN enhanced model, obtain corresponding simulation data based on the generator, and input the simulation data and the first power data as input data into a discriminator to obtain second power data similar to the first power data, wherein the GAN enhanced model includes the generator and the discriminator;

[0050] The control module is used to input the second power data into a preset dynamic optimization strategy network to obtain an optimal control strategy for the power system, and control the power system based on the optimal control strategy.

[0051] The embodiment of the present application can remove noise or outliers that may be contained in the historical operation data by performing data preprocessing on the collected historical operation data, so as to facilitate the subsequent GAN enhanced model to accurately obtain the second power data that is similar to the first power data and of higher quality; after learning the first power data through the generator in the GAN enhanced model, simulated data is generated, and the simulated data and the first power data are distinguished through the discriminator, so as to facilitate the second power data that is similar to the first power data and of higher quality, which can reduce the impact of uncertainty data on subsequent stability analysis, and can provide more reliable input data for subsequent steady-state operation analysis and optimal control strategy, thereby ensuring the stable operation of the power system after the new energy is connected; through the preset dynamic optimization strategy network, the steady-state operation of the power system can be accurately and efficiently evaluated based on the second power data, and the control strategy that is most suitable for the second power data can be generated, thereby ensuring the steady-state operation of the power system after the new energy is connected. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flow chart of an embodiment of the uncertainty data driven steady-state operation control method of a power system provided by the present application;

[0053] Figure 2 It is a structural schematic diagram of an embodiment of the uncertainty data driven steady-state operation control system of the power system provided in the present application. DETAILED DESCRIPTION

[0054] 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.

[0055] It should be understood that the step numbers used in this article are only for the convenience of description and are not intended to limit the order in which the steps are executed.

[0056] It should be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, the singular forms "a", "an" and "the" are intended to include plural forms unless the context clearly indicates otherwise.

[0057] The terms “include” and “comprising” indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0058] The term "and / or" means and includes any and all possible combinations of one or more of the associated listed items.

[0059] As new energy generation technologies are rapidly integrated into the power grid, traditional power systems are facing new challenges. This type of new energy generation relies on natural conditions, and therefore exhibits greater volatility and discontinuity, making it difficult to accurately estimate the output power of the power generation, resulting in a significant increase in the complexity of the power system's dispatch, and also posing a severe challenge to the system's stable operation and overall safety. Traditional steady-state analysis is based on deterministic models, but the integration of new energy into the grid will cause dynamic changes in parameters, bringing uncertainty to system operation. Therefore, traditional means are difficult to predict system status and risks, and thus difficult to deal with a series of problems such as grid frequency fluctuations, voltage over-limits, and even system collapse.

[0060] Next, the nouns involved in this application are analyzed:

[0061] The GAN (Generative Adversarial Networks) enhancement model is a technology that uses GAN for data enhancement. It aims to improve the performance of deep learning models and consists of two key parts: the generator and the discriminator. The goal of the generator is to generate new data that is similar to the original data, while the goal of the discriminator is to distinguish the data generated by the generator from the original data. Through adversarial means, the generator and the discriminator gradually reach a balance during the training process, and the data generated by the generator will become closer and closer to the original data.

[0062] Wavelet transform is a new mathematical analysis method, which is widely used in many fields such as signal processing, graphics processing, pattern recognition, language analysis, etc. It has the ability of localized analysis in the time and frequency domain, with high frequency resolution and low time resolution in the low frequency band, and high time resolution and low frequency resolution in the high frequency band, and can analyze any details of the signal.

[0063] Empirical Mode Decomposition (EMD) is a signal processing technique used to decompose complex nonlinear or non-stationary signals into intrinsic mode functions (IMFs). EMD constructs envelopes by finding local maxima and minima, extracts IMFs, and adaptively adapts to various signal analyses.

[0064] Based on this, the embodiments of the present application provide an uncertainty data-driven steady-state operation control method and system for the power system, which can accurately and efficiently evaluate the steady-state operation of the power system and ensure the steady-state operation of the power system after the access of new energy.

[0065] The uncertainty data driven steady-state operation control method and system for a power system provided in the embodiments of the present application are specifically illustrated through the following embodiments. First, the uncertainty data driven steady-state operation control method for a power system in the embodiments of the present application is described.

[0066] The uncertainty data-driven steady-state operation control method of the power system provided in the embodiment of the present application relates to the field of power system communication. The uncertainty data-driven steady-state operation control method of the power system provided in the embodiment of the present application can be applied to the terminal, can also be applied to the server side, and can also be software running in the terminal or the server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or it can be configured as a server cluster or a distributed system composed of multiple physical servers, and can also be configured as a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application that implements the uncertainty data-driven steady-state operation control method of the power system, but is not limited to the above forms.

[0067] The present application can also be used in numerous general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0068] Embodiment 1

[0069] Please refer to Figure 1 , Figure 1 It is a flowchart of an embodiment of the uncertainty data driven power system steady-state operation control method provided by the present application, including steps S101 to S103;

[0070] Step S101, collecting historical operation data of the new energy power system, and performing data preprocessing on the historical operation data to obtain first power data, wherein the historical operation data includes voltage, frequency and power;

[0071] It is understandable that various sensors can be installed in the new energy power generation system to collect historical operation data of the new energy power system, wherein the power includes active power and reactive power. Since the collected data may contain noise or outliers, it is necessary to preprocess the historical operation data for data cleaning to remove outliers and noise, wherein the means of data preprocessing include but are not limited to wavelet transform, empirical mode decomposition (EMD) and other methods. Specifically, the process of data preprocessing by wavelet transformation is as follows: first, the historical operation data is subjected to wavelet transform to obtain sub-signals of different frequencies, wherein the sub-signals include low-frequency signals and high-frequency signals; secondly, it is determined whether the high-frequency signal is higher than a preset threshold; thirdly, if the high-frequency signal is higher than the preset threshold, the first high-frequency signal higher than the preset threshold is removed, and if the high-frequency signal is lower than the preset threshold, the second high-frequency signal less than the preset threshold is assigned to zero or soft threshold processing to obtain a third high-frequency signal; finally, the third high-frequency signal is reconstructed with the low-frequency signal to obtain the first power data. The process of data preprocessing through empirical mode decomposition (EMD) is specifically as follows: first, the historical operation data is subjected to empirical mode decomposition to obtain a first IMF component and a residual signal; second, it is determined whether the first IMF component meets a preset condition. If the preset condition is met, the second IMF component that meets the preset condition is eliminated to obtain a plurality of third IMF components; finally, the remaining plurality of the third IMF components are added together to obtain the first power data.

[0072] In this way, by preprocessing the collected historical operation data through wavelet transform or empirical mode decomposition, noise or outliers that may be contained in the historical operation data can be removed, making it easier for the subsequent GAN enhanced model to accurately obtain second power data that is similar to the first power data but of higher quality.

[0073] Step S102: inputting the first power data into a preset GAN enhanced model, obtaining corresponding simulation data based on a generator, and inputting the simulation data and the first power data into a discriminator as input data to obtain second power data similar to the first power data, wherein the GAN enhanced model includes the generator and the discriminator;

[0074] It is understandable that after obtaining the cleaned first power data, it can be input into the GAN enhanced model, wherein the GAN enhanced model includes a generator and a discriminator, and combines a variational autoencoder (VAE) and an attention mechanism. After the GAN enhanced model obtains the first power data, it will generate simulated data corresponding to the first power data based on the generator. Specifically: first, the first power data needs to be standardized to obtain the third power data; second, the third power data is encoded by the variational autoencoder VAE to obtain the corresponding latent variable Z VAE ; Finally, the generator G is based on the latent variable Z VAE and random noise z, and uses the introduced attention mechanism to adaptively focus on the key fluctuation patterns of renewable energy generation from the input features to obtain the simulated data X corresponding to the first power data fake , where Z VAE ~q(z|X)=N(μ(X),∑(X)), where Z VAE is a latent variable, μ(X) is the mean of the first power data, and ∑(X) is the variance of the first power data.

[0075] It should be noted that latent variables have the following characteristics: (1) Low-dimensional representation: Latent variables compress high-dimensional input data into low-dimensional representation, so that the model can effectively capture the main features of the data while reducing computational complexity. (2) Distribution characteristics: Latent variables are usually assumed to obey a certain prior distribution (such as a normal distribution). During the training process, VAE optimizes the distribution of latent variables by maximizing the combined loss of likelihood estimation and KL divergence to make it as close to the prior distribution as possible, which helps to generate new samples that are more consistent with the distribution of real data. (3) Information expression: Latent variables reflect the structure and characteristics of the input data to a certain extent, and can capture important information and patterns in the data. In this way, VAE can generate new data samples with similar characteristics. (4) Generation ability: The characteristics of latent variables enable the generator to generate simulated data similar to the distribution of real data based on these variables, thereby improving the quality and accuracy of generated data.

[0076] In this way, the potential distribution characteristics of the first power data are captured by the variational autoencoder, thereby enhancing the initial representation ability of the generator to obtain latent variables, so that the generator can generate simulated data similar to the real data distribution based on these variables, thereby improving the quality and accuracy of the generated data.

[0077] When the simulation data X corresponding to the first power data is obtained fake After that, the simulated data X fake With the first power data X realAs input data, the discriminator D is input, and the first power data X is obtained. real Similar second power data, specifically: first, the simulation data X fake With the first power data X real Scaling or segmenting at different scales is performed to obtain corresponding eigenvalues; secondly, the discriminator discriminates the eigenvalues ​​at each scale to obtain the discrimination probability at the scale; finally, the discrimination probabilities at all scales are comprehensively evaluated to obtain the final discrimination probability, and the second power data similar to the first power data is determined based on the final discrimination probability, wherein the objective function of the discriminator is specifically:

[0078]

[0079] Where, X real is the real first power data, D(X real ) is the discrimination probability of the first power data, X fake is the simulation data, D(X fake ) is the discrimination probability of the simulated data, p data (X) is the actual distribution of the first power data, p G (X) is the distribution of the simulation data.

[0080] It should be noted that the output of the discriminator is the discriminant probability, that is, the output is a discriminant probability between 0 and 1, indicating the possibility that the input data is real data. A value close to 1 indicates that the input data is more likely to be real data, while a value close to 0 indicates that it is generated data.

[0081] It should be noted that in the process of training the GAN enhancement model, after the discriminator generates the discrimination probability at the corresponding scale, it is necessary to optimize the generator and the discriminator through the Wasserstein distance loss function until the generated data can approach the distribution of the actual renewable energy power generation.

[0082] In this way, the discriminator performs multiple discriminations through feature representations at different scales, ensuring that the generator can generate data distributions close to the real one at different time scales, and then obtain second power data similar to the first power data, which can provide more reliable input data for subsequent steady-state operation analysis and optimal control strategies, thereby ensuring the stable operation of the power system after the access of new energy.

[0083] Step S103: input the second power data into a preset dynamic optimization strategy network to obtain an optimal control strategy for the power system, and control the power system based on the optimal control strategy.

[0084] It can be understood that after obtaining the second power data of higher quality and closer to the actual situation through the GAN enhanced model, the second power data needs to be input into the dynamic optimization strategy network to obtain the optimal control strategy of the power system. Specifically: first, based on the second power data, the first control strategy of the power system is determined; secondly, the stability operation state of the new energy power system after the implementation of the first control strategy is analyzed through the dynamic optimization strategy network to obtain the stability evaluation result; finally, based on the stability evaluation result, the control strategy is adjusted through the dynamic optimization strategy network to determine the optimal control strategy of the power system. Among them, by optimizing the objective function, the control strategy θ is determined with the goal of minimizing the deviation of voltage and frequency and the power imbalance. Among them, the objective function of the dynamic optimization strategy network is;

[0085]

[0086] Where, θ is the control strategy; V i and F i are the voltage and frequency of the power system at time t; V ref is the voltage reference value of the power system; F ref is the frequency reference value of the power system; P imbalance (t) is the power imbalance of the power system at time t; α is a control parameter, which represents the weight of the power imbalance.

[0087] It should be noted that the optimal control strategy is usually manifested as the selection of control actions, such as adjusting the output of generators, adjusting loads, or controlling the operation of other equipment in the power grid.

[0088] In this way, the preset dynamic optimization strategy network can accurately and efficiently evaluate the steady-state operation of the power system based on the second power data, generate the control strategy that best suits the second power data, and thus ensure the steady-state operation of the power system after the new energy is connected.

[0089] It should be noted that the dynamic optimization strategy network is achieved through the deep Q network (DQN) algorithm in reinforcement learning. By training the dynamic optimization network, it can ensure that it can autonomously learn the optimal control strategy under the conditions of renewable energy power generation fluctuations, thereby ensuring the efficiency and reliability of steady-state operation analysis.

[0090] Taking into account the real-time volatility of renewable energy power generation, it is necessary to monitor the power operation of the renewable energy power generation system in real time to improve the real-time response capability of the power system. Specifically: first, the operating data of the renewable energy power system acquired in real time is analyzed by the edge computing node to obtain the analysis results; then, the edge computing node dynamically adjusts the control strategy based on the analysis results, and sends the adjusted second control strategy to the power grid dispatching center, so that the power grid dispatching center can determine and execute the third control strategy.

[0091] In this way, the real-time operation data of the power system is received and analyzed by the edge computing nodes in real time, and fed back to the grid dispatching center to ensure that the power system can respond quickly when the renewable energy fluctuates violently, thereby maintaining the steady-state operation of the renewable energy power system.

[0092] The embodiment of the present application can remove noise or outliers that may be contained in the historical operation data by performing data preprocessing on the collected historical operation data, so as to facilitate the subsequent GAN enhanced model to accurately obtain the second power data that is similar to the first power data and of higher quality; after learning the first power data through the generator in the GAN enhanced model, simulated data is generated, and the simulated data and the first power data are distinguished through the discriminator, so as to facilitate the second power data that is similar to the first power data and of higher quality, which can reduce the impact of uncertainty data on subsequent stability analysis, and can provide more reliable input data for subsequent steady-state operation analysis and optimal control strategy, thereby ensuring the stable operation of the power system after the new energy is connected; through the preset dynamic optimization strategy network, the steady-state operation of the power system can be accurately and efficiently evaluated based on the second power data, and the control strategy that is most suitable for the second power data can be generated, thereby ensuring the steady-state operation of the power system after the new energy is connected.

[0093] Embodiment 2

[0094] Please refer to Figure 2 , Figure 2 It is a structural diagram of an embodiment of the uncertainty data driven steady-state operation control system of the power system provided in the present application; it includes: a collection module 100, a processing module 200 and a control module 300;

[0095] The acquisition module 100 is used to acquire historical operation data of the new energy power system and perform data preprocessing on the historical operation data to obtain first power data, wherein the historical operation data includes voltage, frequency and power;

[0096] The processing module 200 is used to input the first power data into a preset GAN enhanced model, obtain corresponding simulation data based on the generator, and input the simulation data and the first power data as input data into a discriminator to obtain second power data similar to the first power data, wherein the GAN enhanced model includes the generator and the discriminator;

[0097] The control module 300 is used to input the second power data into a preset dynamic optimization strategy network to obtain an optimal control strategy for the power system, and control the power system based on the optimal control strategy.

[0098] The information interaction, execution process, etc. between the modules within the above-mentioned uncertainty data driven power system steady-state operation control system are based on the same concept as the embodiment of the uncertainty data driven power system steady-state operation control method of the first aspect of the present invention, and the technical effects achieved are basically the same. For specific contents, please refer to the description in Embodiment 1 of the method of the present invention, and will not be repeated here.

[0099] The device embodiments described above are merely illustrative, wherein the modules described as separate components may or may not be physically separated, i.e., may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the method of this embodiment.

[0100] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the uncertainty data-driven steady-state operation control method of the power system as described in the first embodiment above is implemented.

[0101] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-monitorable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0102] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present application in detail. It should be understood that the above description is only a specific embodiment of the present application and is not intended to limit the scope of protection of the present application.

[0103] It is particularly pointed out that for those skilled in the art, any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for controlling steady-state operation of a power system driven by uncertainty data, characterized in that: include: Collecting historical operation data of the new energy power system and performing data preprocessing on the historical operation data to obtain first power data, wherein the historical operation data includes voltage, frequency and power; Inputting the first power data into a preset GAN enhanced model, obtaining corresponding simulation data based on a generator, and inputting the simulation data and the first power data into a discriminator as input data to obtain second power data similar to the first power data, wherein the GAN enhanced model includes the generator and the discriminator; The second power data is input into a preset dynamic optimization strategy network to obtain an optimal control strategy for the power system, and the power system is controlled based on the optimal control strategy.

2. The uncertainty data driven steady-state operation control method of a power system according to claim 1, characterized in that: The data preprocessing of the historical operation data to obtain the first power data is specifically: Performing wavelet transform on the historical operation data to obtain sub-signals of different frequencies, wherein the sub-signals include low-frequency signals and high-frequency signals; Determining whether the high-frequency signal is higher than a preset threshold; If the high-frequency signal is higher than a preset threshold, the first high-frequency signal higher than the preset threshold is removed; if the high-frequency signal is lower than the preset threshold, the second high-frequency signal lower than the preset threshold is assigned a value of zero or soft threshold processing to obtain a third high-frequency signal; The third high-frequency signal and the low-frequency signal are reconstructed to obtain the first power data.

3. The uncertainty data driven steady-state operation control method of a power system according to claim 1, characterized in that: The data preprocessing of the historical operation data to obtain the first power data is specifically: Performing empirical mode decomposition on the historical operation data to obtain a first IMF component and a residual signal; Determine whether the first IMF component meets a preset condition, and if so, remove the second IMF component that meets the preset condition to obtain a plurality of third IMF components; The remaining several third IMF components are added together to obtain the first power data.

4. The uncertainty data driven steady-state operation control method of a power system according to claim 1, characterized in that: The first power data is input into a preset GAN enhancement model, and corresponding simulation data is obtained based on the generator, specifically: performing standardization processing on the first power data to obtain third power data; Encoding the third power data by a variational autoencoder to obtain corresponding latent variables; The generator obtains simulation data corresponding to the first power data based on the latent variables and random noise.

5. The uncertainty data driven steady-state operation control method of a power system according to claim 1, characterized in that: The analog data and the first power data are input into a discriminator as input data to obtain second power data similar to the first power data, specifically: Scaling or segmenting the simulation data and the first power data at different scales to obtain corresponding eigenvalues; The discriminator discriminates the feature value at each scale to obtain a discriminant probability at the scale; The discrimination probabilities at all scales are comprehensively evaluated to obtain a final discrimination probability, and second power data similar to the first power data is determined based on the final discrimination probability.

6. The uncertainty data driven steady-state operation control method for a power system according to any one of claim 1, characterized in that: The objective function of the discriminator is specifically: Where, X real is the real first power data, D(X real ) is the discrimination probability of the first power data, X fake is the simulation data, D(X fake ) is the discrimination probability of the simulated data, p data (X) is the actual distribution of the first power data, p G (X) is the distribution of the simulation data.

7. The uncertainty data driven steady-state operation control method of a power system according to claim 1, characterized in that: The step of inputting the second power data into a preset dynamic optimization strategy network to obtain an optimal control strategy for the power system is specifically as follows: determining a first control strategy for the power system based on the second power data; The stability evaluation result is obtained by analyzing the stability operation state of the new energy power system after the first control strategy is implemented through the dynamic optimization strategy network; Based on the stability evaluation result, the control strategy is adjusted through the dynamic optimization strategy network to determine the optimal control strategy for the power system.

8. The uncertainty data driven steady-state operation control method of a power system according to claim 1, characterized in that: The objective function of the dynamic optimization strategy network is: Where, θ is the control strategy; V i and F i are the voltage and frequency of the power system at time t; V ref is the voltage reference value of the power system; F ref is the frequency reference value of the power system; P imbalance (t) is the power imbalance of the power system at time t; α is a control parameter, which represents the weight of the power imbalance.

9. The uncertainty data driven steady-state operation control method for a power system according to any one of claims 1 to 8, characterized in that: After the power system is controlled based on the optimal control strategy, the method further includes: The operating data of the new energy power system acquired in real time is analyzed through the edge computing node to obtain the analysis results; The edge computing node dynamically adjusts the control strategy based on the analysis result, and sends the adjusted second control strategy to the power grid dispatching center, so that the power grid dispatching center determines and executes the third control strategy.

10. An uncertainty data driven power system steady-state operation control system, characterized in that: include: Acquisition module, processing module and control module; The acquisition module is used to collect historical operation data of the new energy power system and perform data preprocessing on the historical operation data to obtain first power data, wherein the historical operation data includes voltage, frequency and power; The processing module is used to input the first power data into a preset GAN enhanced model, obtain corresponding simulation data based on the generator, and input the simulation data and the first power data as input data into a discriminator to obtain second power data similar to the first power data, wherein the GAN enhanced model includes the generator and the discriminator; The control module is used to input the second power data into a preset dynamic optimization strategy network to obtain an optimal control strategy for the power system, and control the power system based on the optimal control strategy.

Images