Modeling method and device of power system, computer program product and modeling system

The intelligent generative model is constructed through neural networks and numerical integral algorithms, which solves the problem of poor accuracy of the power system model and improves analysis efficiency and accuracy, especially in small disturbance stability analysis.

CN119026472APending Publication Date: 2024-11-26ELECTRIC POWER RES INST CHINA SOUTHERN POWER GRID CO LTD +1
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411116111.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-14
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The accuracy of the power system model constructed in the prior art is poor, resulting in low efficiency in the stability analysis of the power system.

Method used

An intelligent generative model is constructed using neural network algorithms and numerical integral algorithms, and a virtual model is generated through multiple sets of training data to simulate the dynamic process of the power system.

Benefits of technology

The accuracy and analysis efficiency of the power system model are improved, and the accuracy and interpretability of small disturbance stability analysis are enhanced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119026472A_ABST
    Figure CN119026472A_ABST
Patent Text Reader

Abstract

The invention provides a modeling method and device of a power system, a computer program product and a modeling system. The method comprises the following steps: acquiring related information of a power system; constructing an intelligent generation model; inputting the related information into the intelligent generation model to obtain virtual model data corresponding to the related information; modeling is carried out according to the virtual model data corresponding to the related information, a virtual model is obtained, and the virtual model is a virtual model of the power system. According to the scheme, the problem that the accuracy of the constructed model of the power system in the prior art is relatively poor is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of intelligent power systems, and in particular to a modeling method, device, computer program product and modeling system for a power system. Background Art

[0002] In the current development of power systems, we are faced with complex dynamic and coupling characteristics introduced by a high proportion of renewable energy and power electronic devices, which brings new challenges to the stability analysis and control of the system.

[0003] Currently, for the power system, if stability analysis is to be carried out, modeling is required. However, in the current solution, the accuracy of the constructed power system model is poor, resulting in low efficiency of subsequent stability analysis of the power system. Summary of the invention

[0004] The main purpose of the present application is to provide a method, device, computer program product and modeling system for a power system, so as to at least solve the problem of poor accuracy of the power system model constructed in the prior art.

[0005] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a modeling method for an electric power system is provided, comprising: obtaining relevant information of the electric power system, wherein the relevant information is operating information of electric power equipment, wherein the electric power equipment includes at least one or more of power generation equipment, transmission equipment, distribution equipment and power consumption equipment, and the operating information includes one or more of phase angle, rotation speed, transient potential, voltage, current, power and frequency; constructing an intelligent generation model, wherein the intelligent generation model is obtained by training with a neural network algorithm and a numerical integration algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes historical relevant information obtained within a historical time period and historical virtual model data corresponding to the historical relevant information, wherein the historical virtual model data is virtual model data corresponding to the historical electric power system; inputting the relevant information into the intelligent generation model to obtain virtual model data corresponding to the relevant information; modeling according to the virtual model data corresponding to the relevant information to obtain the virtual model, wherein the virtual model is a virtual model of the electric power system.

[0006] Optionally, before constructing the intelligent generation model, the method further includes: obtaining a differential algebraic equation, wherein the differential algebraic equation is:

[0007]

[0008] x represents the state quantity of the power system, represents the derivative of the state quantity, y represents the algebraic quantity, f() represents the differential equation part describing the power system, and g() represents the algebraic equation part describing the power system; the non-state quantities in the differential algebraic equation are removed to obtain a differential equation group, wherein the differential equation group is: The differential equation group is used to construct the intelligent generation model.

[0009] Optionally, before constructing the intelligent generation model, the method further includes: calculating the state quantity of the power system at the target time according to a first formula, wherein the first formula is:

[0010]

[0011] t k represents the target time, x(t k ) represents the state quantity of the power system at the target time, t0 represents the initial time, x(t0) represents the state quantity of the power system at the initial time, f θ (x)=θ1x+θ2, θ1 represents the weight coefficient, and θ2 represents the bias coefficient.

[0012] Optionally, before constructing the intelligent generation model, the method further includes: using a sequence expansion formula to expand the sample data to obtain expanded sample data, wherein the sequence expansion formula is:

[0013] y e,t,i =θ 3,i ·[y o,t-r,i ,y o,t-r+1,i ,...,y o,t+r-1,i ,y o,t+r,i ] T ,

[0014] y e,t,i represents the value of the i-th supplementary quantity measured at time t obtained by expansion, θ 3,i represents the network weight parameter of the i-th quantity measurement, y o,t,i represents the value of the i-th element of the original quantity measurement at time t, t represents the target time, r represents the time window length of each sequence expander, and T represents the matrix transpose symbol.

[0015] Optionally, before constructing the intelligent generation model, the method also includes: constructing an initial intelligent generation model, wherein the initial intelligent generation model is trained by a neural network algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes the historical relevant information obtained within a historical time period and the historical virtual model data corresponding to the historical relevant information; using numerical integration and mean square error loss function to optimize the initial intelligent generation model to obtain the intelligent generation model.

[0016] Optionally, after modeling the virtual model data corresponding to the relevant information to obtain the virtual model, the method further includes: obtaining the similarity between the virtual model and the power system; determining the accuracy of the intelligent generation model based on the similarity, wherein the similarity and the accuracy are positively correlated.

[0017] Optionally, before obtaining relevant information of the power system, the method also includes: obtaining initial relevant information of the power system, wherein the initial relevant information is initial operating information of the power equipment; preprocessing the initial relevant information to obtain the relevant information, wherein the data preprocessing method includes one or more of denoising, standardization and normalization.

[0018] According to another aspect of the present application, a modeling device for an electric power system is provided, comprising: a first acquisition unit, for acquiring relevant information of the electric power system, wherein the relevant information is operating information of electric power equipment, wherein the electric power equipment includes at least one or more of power generation equipment, transmission equipment, distribution equipment and power consumption equipment, and the operating information includes one or more of phase angle, rotation speed, transient potential, voltage, current, power and frequency; a first construction unit, for constructing an intelligent generation model, wherein the intelligent generation model is obtained by training with a neural network algorithm and a numerical integration algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes historical relevant information obtained within a historical time period and historical virtual model data corresponding to the historical relevant information, wherein the historical virtual model data is virtual model data corresponding to the historical electric power system; a first processing unit, for inputting the relevant information into the intelligent generation model to obtain virtual model data corresponding to the relevant information; a second processing unit, for modeling according to the virtual model data corresponding to the relevant information to obtain the virtual model, wherein the virtual model is a virtual model of the electric power system.

[0019] According to another aspect of the present application, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the steps of any one of the power system modeling methods are implemented.

[0020] According to another aspect of the present application, a modeling system is provided, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing any one of the modeling methods of the power system described.

[0021] By applying the technical solution of the present application, a model can be constructed by using a neural network algorithm and a numerical integration algorithm. The neural network algorithm can learn a large amount of data for rapid learning, while the numerical integration algorithm can accurately simulate various dynamic processes in the power system. Therefore, the intelligent generation model constructed by this solution has high accuracy, and the accuracy of the obtained power system model is good. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings constituting part of the present application are used to provide a further understanding of the present application. The exemplary embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0023] Figure 1 A hardware structure block diagram of a mobile terminal for executing a modeling method for a power system provided in an embodiment of the present application is shown;

[0024] Figure 2 A schematic diagram of a flow chart of a method for modeling a power system provided according to an embodiment of the present application is shown;

[0025] Figure 3 A flowchart of the power system data-driven small disturbance modeling method showing the interpretability of the present scheme;

[0026] Figure 4 A structural block diagram of a power system modeling device provided according to an embodiment of the present application is shown.

[0027] The above drawings include the following reference numerals:

[0028] 102, processor; 104, memory; 106, transmission device; 108, input and output devices. DETAILED DESCRIPTION

[0029] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0030] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings 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 should fall within the scope of protection of the present application.

[0031] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] In the current development of power systems, we are faced with complex dynamic and coupling characteristics introduced by a high proportion of renewable energy and power electronic devices, which brings new challenges to the stability analysis and control of the system. Although traditional small perturbation stability analysis methods, such as eigenvalue analysis and modal identification, have been widely used, they are insufficient in accuracy and comprehensiveness in new power systems. In recent years, although artificial intelligence methods have made progress in stability analysis, they usually lack detailed understanding and interpretability of system dynamics. Therefore, it is particularly important to develop a new small perturbation modeling method that can analyze system dynamics in detail and has interpretability.

[0033] As introduced in the background technology, the accuracy of the power system model constructed in the prior art is poor. To solve the above problems, the embodiments of the present application provide a power system modeling method, device, computer program product and modeling system.

[0034] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0035] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 FIG. 1 is a hardware structure block diagram of a mobile terminal of a power system modeling method according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0036] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the display method of device information in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof. The transmission device 106 is used to receive or send data via a network. The above-mentioned specific examples of the network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0037] In this embodiment, a method for modeling an electric power system running on a mobile terminal, a computer terminal or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0038] Figure 2 FIG. 1 is a flow chart of a method for modeling a power system according to an embodiment of the present application. Figure 2 As shown, the method comprises the following steps:

[0039] Step S201, obtaining relevant information of the power system, wherein the relevant information is operation information of the power equipment, wherein the power equipment at least includes one or more of power generation equipment, power transmission equipment, power distribution equipment and power consumption equipment, and the operation information includes one or more of phase angle, speed, transient potential, voltage, current, power and frequency;

[0040] Specifically, the power system is a complex network consisting of four main links: power generation, transmission, distribution, and consumption, which is responsible for the generation, transmission, distribution, and use of electric energy. Power generation equipment such as generators are responsible for generating electric energy; transmission equipment such as transmission lines and transformers are responsible for transmitting electric energy from power plants to various places; distribution equipment is responsible for distributing electric energy to various users; and power consumption equipment is the equipment that ultimately uses electric energy, such as household appliances, industrial equipment, etc. This information involves single or multiple types of equipment.

[0041] In an AC system, the phase angle reflects the phase position of a voltage or current waveform relative to a reference point (such as zero). Speed: refers specifically to the rotation speed of a rotating device such as a generator. For a generator, the speed is directly related to the frequency of the electrical energy emitted. Transient potential: When a fault or disturbance occurs in the power system, the instantaneous value of the equipment potential is crucial for analyzing system stability and troubleshooting. Voltage: The basic parameter for the transmission and distribution of electrical energy in a power system, which indicates the ability of the electric field force to do work. Current: A physical quantity formed by the directional movement of electric charges, which reflects the flow of electrical energy. Power: A measure of the electrical energy converted, used, or dissipated per unit time, which is an important indicator for evaluating equipment performance and system load. Frequency: The number of periodic changes in AC power per unit time determines the rate of periodic changes in electrical energy.

[0042] Of course, in addition to the above information, other relevant information can also be collected.

[0043] Step S202, constructing an intelligent generation model, wherein the intelligent generation model is obtained by training with a plurality of sets of training data through a neural network algorithm and a numerical integration algorithm, wherein each set of training data in the plurality of sets of training data includes historical related information acquired in a historical time period and historical virtual model data corresponding to the historical related information, wherein the historical virtual model data is virtual model data corresponding to a historical power system;

[0044] Specifically, machine learning training can be performed through a neural network algorithm and a numerical integration algorithm to obtain an intelligent generation model, and through the intelligent generation model, relevant data for constructing a virtual model of the power system can be obtained.

[0045] Neural networks are often used to extract useful features from input data and make preliminary predictions or classifications. In power systems, inputs may include historical load data, weather conditions, holiday information, etc. Neural networks can learn the complex relationships between these inputs and outputs (such as future loads).

[0046] Output as initial conditions for numerical integration: The output of a neural network can be used as initial conditions or input parameters for a numerical integration algorithm. For example, if a neural network predicts the state of a system (such as voltage, current, etc.) at a certain point in the future, these predicted values ​​can be used as the starting point for a numerical integration algorithm to simulate the dynamic behavior of the system at a certain time in the future.

[0047] Numerical integration algorithms (such as Euler method, Runge-Kutta method, etc.) are used to simulate the dynamic behavior of power systems. These algorithms can gradually calculate the state of the system at future time points based on the physical equations of the system (such as circuit equations, dynamic equations, etc.) and initial conditions (provided by neural networks).

[0048] Feedback and Adjustment: In some cases, the results of numerical integration can be fed back to the neural network to adjust the parameters or weights of the neural network. This feedback mechanism can make the model more accurate in predicting the future behavior of the system.

[0049] In the training phase, the neural network and the numerical integration algorithm can be trained as a whole. By comparing the difference between the model's prediction results and the actual observed data, the weights of the neural network and the parameters of the numerical integration algorithm are adjusted using an optimization algorithm (such as the gradient descent method).

[0050] Alternatively, the neural network can be trained separately to enable it to accurately predict the initial state of the system or certain key parameters; then the trained neural network can be combined with the numerical integration algorithm for overall simulation and prediction.

[0051] Step S203, inputting the above-mentioned relevant information into the above-mentioned intelligent generation model to obtain virtual model data corresponding to the above-mentioned relevant information;

[0052] Specifically, the relevant information collected above is passed to the intelligent generation model that has been constructed. The model will process and analyze the information according to its internal structure and algorithm, and generate corresponding output data.

[0053] Step S204, modeling is performed according to the virtual model data corresponding to the relevant information to obtain the virtual model, wherein the virtual model is a virtual model of the power system.

[0054] Specifically, the virtual model is modeled according to the virtual model data to obtain a virtual model, which is a digital representation of the power system. It contains information such as the structure, parameters, and operation rules of the power system. The virtual model can simulate various operating states and behaviors of the power system, such as load changes, fault occurrences, and dispatch operations, and generate corresponding simulation data. These simulation data can be used for analysis, prediction, optimization, and decision support of the power system.

[0055] Through this embodiment, a model can be constructed by using a neural network algorithm and a numerical integration algorithm. The neural network algorithm can learn a large amount of data for rapid learning, while the numerical integration algorithm can accurately simulate various dynamic processes in the power system. Therefore, the intelligent generation model constructed by this scheme has a high accuracy, and the accuracy of the obtained power system model is good.

[0056] Specifically, the solution of the present application establishes a linear network neural differential equation model (i.e., intelligent generation model) with clear physical meaning to simulate the small disturbance behavior of the power system. In view of the problem that some state quantities in the actual power system are difficult to measure directly, an improved model structure with order expansion of adjacent time series sampling points is adopted to improve the applicability of the model. A data-driven small disturbance stable neural differential equation model training method is implemented, and the time domain small disturbance waveform data of the power system is used for model training.

[0057] Specifically, the solution of this application proposes a new method for small-disturbance stability analysis of power systems by combining advanced models of neural differential equations and data-driven training methods. This method not only improves the accuracy of small-disturbance stability analysis, but also enhances the interpretability of analysis results due to its modeling method with clear physical meaning, effectively improving the efficiency and accuracy of small-disturbance analysis of power systems. This solution takes into account the convenience of user operation and the efficiency of model training, and effectively improves the overall performance of small-disturbance analysis of power systems.

[0058] In the specific implementation process, before constructing the intelligent generation model, the method further includes the following steps: obtaining a differential algebraic equation, wherein the differential algebraic equation is:

[0059]

[0060] x represents the state quantity of the above power system, represents the derivative of the above state quantity, y represents the algebraic quantity, f() represents the differential equation part describing the above power system, and g() represents the algebraic equation part describing the above power system; the non-state quantity in the above differential algebraic equation is removed to obtain a differential equation group, wherein the above differential equation group is: The above differential equations are used to construct the above intelligent generation model.

[0061] In this scheme, a linear network neural differential equation model is created to approximate the dynamic characteristics of the power system. For the establishment of a small disturbance model (intelligent generation model) of the power system, this scheme takes into account the interaction and dynamic behavior of various electrical components in the power system. The dynamic behavior of the power system can be described by a set of differential algebraic equations, which are in the form of differential algebraic equations. According to the implicit function theorem, the non-state quantities can be eliminated according to the function g, so that the differential algebraic equation group can be merged into a differential equation group only about the state quantity. This scheme can create a linear network neural differential equation model. Considering the complexity of the power system, advanced neural network technology is used to approximate the dynamic characteristics of the power system. In the process of building the model, various dynamic factors in the power system are taken into account, such as the phase angle and speed of the generator. Neural networks are used to represent the differential equations of the power system, which can better describe the dynamic response of the power system under small disturbances.

[0062] Specifically, the selection of specific state quantities depends on the specific conditions of the target network's equipment, structure, modeling method, etc., and generally may include the generator's phase angle, rotation speed, transient potential, etc.

[0063] For the small disturbance model, due to the small disturbance degree, the system presents an approximate linearization characteristic near the operating point. The linearized system dynamics at the operating point x0 can be expressed as:

[0064]

[0065] A is the system matrix, δx=x-x0.

[0066] This method constructs a model for small perturbation stability analysis. The previous formula is a general formula, generally used for large perturbation analysis. In the specific scenario of small perturbation analysis, this formula only works near a certain working point and presents a linearized state. Here, the differential algebraic equations of the original system are linearized at the working point to obtain a small perturbation model. The equivalent expression obtained by learning the neural differential method later is also this linearized model.

[0067] Specifically, a data-driven neural differential equation model (i.e., intelligent generation model) is established based on the power system model. The effectiveness of using neural differential equations to model physical systems including power systems has been theoretically proven. When the solution step size of the numerical integration process of the neural differential equation is small enough, the neural differential equation can achieve continuous approximation of the original system. Therefore, by generating training data through sufficient sampling of the original system, selecting a suitable network structure, an appropriate numerical integration algorithm and an integration step size, the neural differential equation can achieve continuous approximation of the power system to be modeled, thereby realizing data-driven dynamic modeling of the power system.

[0068] Specifically, this scheme defines the structure and parameters of the model to adapt to the specific dynamic characteristics of the power system; uses a data-driven approach to optimize the neural network model to ensure that the model can accurately capture the small disturbance response of the power system.

[0069] Specifically, this scheme can use neural networks to approximate the differential equations of the power system to capture the key dynamic characteristics of the system; design the model to adapt to different types of power systems to ensure that the model can effectively cope with various small disturbances.

[0070] In the specific implementation process, before constructing the intelligent generation model, the method further includes the following steps: calculating the state quantity of the power system at the target time according to the first formula, wherein the first formula is:

[0071]

[0072] t k represents the above target time, x(t k ) represents the state quantity of the power system at the target time, t0 represents the initial time, x(t0) represents the state quantity of the power system at the initial time, f θ (x)=θ1x+θ2, θ1 represents the weight coefficient, and θ2 represents the bias coefficient.

[0073] In this scheme, neural differential equation technology can be used to learn the timing waveform of the power system, obtain an interpretable differential equation model, and then extract the model information for interpretability analysis, thereby ensuring that the model of this scheme has good interpretability.

[0074] Specifically, according to the basic theory of neural differential equations, the differential equation of the target power system is approximated by the function represented by the neural network, which can be written as: The newly appearing symbol θ in the formula represents the network parameters of the model, which can be adjusted through training. Using the above network, the target system can be predicted through the first formula to obtain the predicted state quantity. It can correspond to the previous differential algebraic equation of the power system. The original equation can be obtained by eliminating the algebraic equation to obtain the dynamics of the power system described by the differential equation. Here is the approximate expression of the equation obtained by neural network training.

[0075] According to the basic theory of neural differential equations, this operation is implemented by an arbitrary numerical integrator, such as the forward Euler algorithm, the implicit Euler algorithm, or the Runge-Kutta algorithm, and the gradient is propagated by adjoint sensitivity or back-propagation methods.

[0076] In the case of small disturbances, the state quantity of the original system presents approximately linear dynamic characteristics near the equilibrium point. Since the system dynamics to be learned only contains linear characteristics, a single-layer perceptron without activation function can be used as the network structure. At this time, the network described by the neural differential equation can be expressed as: θ (x) = θ1x + θ2, θ1 represents the weight parameter of the linear layer of the network, and θ2 represents the bias parameter of the linear layer of the network.

[0077] In the specific implementation process, before building the intelligent generation model, the method further includes the following steps: using a sequence expansion formula to expand the sample data to obtain the expanded sample data, wherein the sequence expansion formula is:

[0078] y e,t,i =θ 3,i ·[y o,t-r,i ,y o,t-r+1,i ,...,y o,t+r-1,i ,y o,t+r,i ] T ,

[0079] y e,t,i represents the value of the i-th supplementary quantity measured at time t obtained by expansion, θ 3,i represents the network weight parameter of the i-th quantity measurement, y o,t,i represents the value of the i-th element of the original quantity measurement at time t, t represents the target time, r represents the time window length of each sequence expander, and T represents the matrix transpose symbol.

[0080] In this scheme, the actual measured data of the power system is processed and the order expansion is performed. The sequence expansion technique is applied to the state quantities that cannot be directly measured to supplement the data set and improve the expressiveness of the model. The structure of the neural differential equation model is improved by the sequence expansion method to enhance the expressiveness of the model under incomplete measurement conditions. This scheme involves introducing sampling points before and after the current moment as model inputs, and its formula can be expressed as a sequence expansion formula. Considering that the state quantities of the power system are often difficult to measure directly in actual scenarios, the order expansion operation is further implemented. This operation involves introducing additional sampling point data to supplement the existing data set, thereby solving the problem of data incompleteness caused by the difficulty of directly measuring the state quantities.

[0081] Specifically, y e,t,iIt can be of order s as needed. This solution can apply a variety of data preprocessing methods to optimize the quality of input data; combined with the specific needs of the power system, the order expansion of the data is performed to make up for the shortcomings of state quantity measurement. For example, when r is 2, the data of the two moments before and after the target moment are used as input. Quantity measurement is the sample data that can be measured. These data may be of insufficient order and face insufficient measurement, which does not meet the observability of the system. The sequence expansion order can be expanded according to the existing state quantity to meet the observability of the system.

[0082] Specifically, the data set can also be constructed and preprocessed. This scheme adopts the following steps: First, use the power system analysis software package (such as PSAT) to build a power system model (i.e., a virtual model) in the Matlab environment, and apply a normal random disturbance with a standard deviation of 0.1% of the steady-state value to each state quantity to simulate a small disturbance event in the actual system. Record the time series change trajectory of the required electrical quantity or state quantity and perform data normalization. Then, divide the sample set into a training set and a test set.

[0083] In some embodiments, before constructing the intelligent generation model, the method further includes the following steps: constructing an initial intelligent generation model, wherein the initial intelligent generation model is trained by a neural network algorithm using multiple sets of training data, and each set of training data in the multiple sets of training data includes the historical relevant information obtained within a historical time period and the historical virtual model data corresponding to the historical relevant information; using numerical integration and mean square error loss function to optimize the initial intelligent generation model to obtain the intelligent generation model.

[0084] In this scheme, numerical integration and optimization algorithms are used to adjust the parameters of the neural network so that the model can more accurately reflect the dynamic changes of the power system under small disturbances. The actual operating data of the power system and artificially simulated small disturbance events are considered in the training process to ensure that the model can accurately predict the behavior of the system under different operating conditions.

[0085] Specifically, during the model training process, a specific numerical integration method and mean square error loss function can be used to optimize the model parameters. The optimization formula used is the second formula, which is an integral formula, namely:

[0086]

[0087] in, represents the state value at time t0, x0 represents the state value at the initial time, h represents the Euler method integration step, and n represents the time from x0 to traversal of the time steps.

[0088] As mentioned above, the model training based on the actual power system operation data is implemented to ensure that the model can accurately reflect the actual situation; advanced numerical integration techniques and optimization strategies are used to adjust and optimize the model parameters. The model parameters are adjusted using numerical integration and optimization algorithms to ensure that the model can accurately reflect the behavior of the power system under small disturbances; the model is trained for multiple rounds, using actual power system data and simulated disturbance events to optimize the model performance. This is the integral formula of the forward Euler method, the value of the system state quantity at time t0 is known By integrating this formula, we can get the value of the state quantity at time t0:

[0089] In some embodiments, after modeling the virtual model data corresponding to the relevant information to obtain the virtual model, the method further includes the following steps: obtaining the similarity between the virtual model and the power system; determining the accuracy of the intelligent generation model based on the similarity, wherein the similarity and the accuracy are positively correlated.

[0090] In this solution, the interpretability of the trained model can be evaluated. The accuracy of the model can be further evaluated by comparing the similarity between the system state predicted by the neural network model and the actual system state. At the same time, the interpretability of the model can be further improved by analyzing the relationship between the neural network model parameters and the system matrix.

[0091] In addition, the trained model can capture the dynamic characteristics of the target system, give an estimate of the derivative function under the current state of the system, and then obtain subsequent time series changes through numerical integration.

[0092] At this point, the network model That is the system dynamic function In the case of small perturbations, the neural network model is represented by f θ (x) = θ1x + θ2, the system dynamics is expressed as = Ax + b = A (x s )(x s is the system operating point), then the network parameters θ1, θ2 and the system matrix θ2 and the equilibrium point x s There is a clear corresponding relationship. Considering that the data preprocessing stage performs data normalization on the sample set, the relationship between the system matrix A of the original system and the estimated value of the bias b and θ1 and θ2 is as follows:

[0093]

[0094] Where S and M are the diagonal matrices of σ' and μ respectively.

[0095] When A is reversible, the system equilibrium point x s The estimated values ​​are:

[0096]

[0097] After obtaining the predicted value of the system A matrix, the traditional small disturbance stability eigenvalue analysis method based on mechanism modeling is also applicable to this model. For example, the eigenvalues ​​and eigenvectors of the system can be obtained through modal decomposition, and then the oscillation modal damping ratio and associated units can be analyzed. A neural network model is learned using time series data, which can give a predicted value that is close to the original power system. The parameters of the model correspond to the state matrix of the original system. However, for the convenience of training, the data is normalized during training and some multiplication and division are performed. Therefore, in order to restore the original value, the extracted matrix needs to be multiplied and divided in reverse to restore it. This process is described here. σ' and μ are factors during normalization, representing the standard deviation and mean.

[0098] In some embodiments, before obtaining relevant information of the power system, the method further includes the following steps: obtaining initial relevant information of the power system, wherein the initial relevant information is initial operation information of the power equipment; preprocessing the initial relevant information to obtain the relevant information, wherein the data preprocessing method includes one or more of denoising, standardization and normalization.

[0099] In this scheme, for the initial measured initial relevant information, the collected data is cleaned and preprocessed, including denoising, standardization, normalization and other operations to improve the data quality to ensure that the data quality meets the input requirements of the model.

[0100] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, the implementation process of the power system modeling method of the present application will be described in detail below in conjunction with specific embodiments.

[0101] This embodiment relates to a specific power system modeling method, that is, an interpretable power system data-driven small disturbance modeling method, such as Figure 3 As shown, the following steps are included:

[0102] Step S1: Establish a power system dynamic model: Use differential algebraic equations to describe various electrical components in the power system and their interactions;

[0103] Step S2: Create a neural differential equation model: According to the power system model, construct and define the structure and parameters of the linear network neural differential equation model;

[0104] Step S3: Data processing and order expansion: Processing the actual measured data of the power system and performing order expansion to expand the incomplete data;

[0105] Step S4: Training the neural differential equation model: The model is trained based on the processed data to optimize the dynamics of small disturbances in the power system.

[0106] The embodiment of the present application also provides a modeling device for an electric power system. It should be noted that the modeling device for an electric power system in the embodiment of the present application can be used to execute the modeling method for an electric power system provided in the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0107] The following is an introduction to the power system modeling device provided in an embodiment of the present application.

[0108] Figure 4 is a structural block diagram of a modeling device for a power system according to an embodiment of the present application. Figure 4 As shown, the device comprises:

[0109] A first acquisition unit 10 is used to acquire relevant information of the power system, wherein the relevant information is operation information of the power equipment, wherein the power equipment at least includes one or more of power generation equipment, power transmission equipment, power distribution equipment and power consumption equipment, and the operation information includes one or more of phase angle, speed, transient potential, voltage, current, power and frequency;

[0110] A first construction unit 20 is used to construct an intelligent generation model, wherein the intelligent generation model is obtained by training with a plurality of sets of training data through a neural network algorithm and a numerical integration algorithm, wherein each set of training data in the plurality of sets of training data includes historical related information obtained within a historical time period and historical virtual model data corresponding to the historical related information, wherein the historical virtual model data is virtual model data corresponding to a historical power system;

[0111] The first processing unit 30 is used to input the above-mentioned relevant information into the above-mentioned intelligent generation model to obtain virtual model data corresponding to the above-mentioned relevant information;

[0112] The second processing unit 40 is used to perform modeling according to the virtual model data corresponding to the relevant information to obtain the virtual model, wherein the virtual model is a virtual model of the power system.

[0113] Through this embodiment, a model can be constructed by using a neural network algorithm and a numerical integration algorithm. The neural network algorithm can learn a large amount of data for rapid learning, while the numerical integration algorithm can accurately simulate various dynamic processes in the power system. Therefore, the intelligent generation model constructed by this scheme has a high accuracy, and the accuracy of the obtained power system model is good.

[0114] In the specific implementation process, the above device also includes a second acquisition unit, which is used to obtain a differential algebraic equation before building the intelligent generation model, wherein the above differential algebraic equation is:

[0115]

[0116] x represents the state quantity of the above power system, represents the derivative of the above state quantity, y represents the algebraic quantity, f() represents the differential equation part describing the above power system, and g() represents the algebraic equation part describing the above power system; the non-state quantity in the above differential algebraic equation is removed to obtain a differential equation group, wherein the above differential equation group is: The above differential equations are used to construct the above intelligent generation model.

[0117] In this scheme, a linear network neural differential equation model is created to approximate the dynamic characteristics of the power system. For the establishment of a small disturbance model (intelligent generation model) of the power system, this scheme takes into account the interaction and dynamic behavior of various electrical components in the power system. The dynamic behavior of the power system can be described by a set of differential algebraic equations, which are in the form of differential algebraic equations. According to the implicit function theorem, the non-state quantities can be eliminated according to the function g, so that the differential algebraic equation group can be merged into a differential equation group only about the state quantity. This scheme can create a linear network neural differential equation model. Considering the complexity of the power system, advanced neural network technology is used to approximate the dynamic characteristics of the power system. In the process of building the model, various dynamic factors in the power system are taken into account, such as the phase angle and speed of the generator. Neural networks are used to represent the differential equations of the power system, which can better describe the dynamic response of the power system under small disturbances.

[0118] In a specific implementation process, the device further includes a first calculation unit, which is used to calculate the state quantity of the power system at the target time according to a first formula before building the intelligent generation model, wherein the first formula is:

[0119]

[0120] t k represents the above target time, x(t k) represents the state quantity of the power system at the target time, t0 represents the initial time, x(t0) represents the state quantity of the power system at the initial time, f θ (x)=θ1x+θ2, θ1 represents the weight coefficient, and θ2 represents the bias coefficient.

[0121] In this scheme, neural differential equation technology can be used to learn the timing waveform of the power system, obtain an interpretable differential equation model, and then extract the model information for interpretability analysis, thereby ensuring that the model of this scheme has good interpretability.

[0122] In the specific implementation process, the above-mentioned device also includes a third processing unit, which is used to expand the sample data using a sequence expansion formula before building the intelligent generation model to obtain the expanded sample data, wherein the above-mentioned sequence expansion formula is:

[0123] y e,t,i =θ 3,i ·[y o,t-r,i ,y o,t-r+1,i ,...,y o,t+r-1,i ,y o,t+r,i ] T ,

[0124] y e,t,i represents the value of the i-th supplementary quantity measured at time t obtained by expansion, θ 3,i represents the network weight parameter of the i-th quantity measurement, y o,t,i represents the value of the i-th element of the original quantity measurement at time t, t represents the target time, r represents the time window length of each sequence expander, and T represents the matrix transpose symbol.

[0125] In this scheme, the actual measured data of the power system is processed and the order expansion is performed. The sequence expansion technique is applied to the state quantities that cannot be directly measured to supplement the data set and improve the expressiveness of the model. The structure of the neural differential equation model is improved by the sequence expansion method to enhance the expressiveness of the model under incomplete measurement conditions. This scheme involves introducing sampling points before and after the current moment as model inputs, and its formula can be expressed as a sequence expansion formula. Considering that the state quantities of the power system are often difficult to measure directly in actual scenarios, the order expansion operation is further implemented. This operation involves introducing additional sampling point data to supplement the existing data set, thereby solving the problem of data incompleteness caused by the difficulty of directly measuring the state quantities.

[0126] In some embodiments, the above-mentioned device also includes a second construction unit and an optimization unit, the second construction unit is used to construct an initial intelligent generation model before constructing the intelligent generation model, wherein the above-mentioned initial intelligent generation model is obtained by training through a neural network algorithm using multiple sets of training data, and each set of training data in the above-mentioned multiple sets of training data includes the above-mentioned historical related information obtained within a historical time period and the above-mentioned historical virtual model data corresponding to the above-mentioned historical related information; the optimization unit is used to use numerical integration and mean square error loss function to optimize the above-mentioned initial intelligent generation model to obtain the above-mentioned intelligent generation model.

[0127] In this scheme, numerical integration and optimization algorithms are used to adjust the parameters of the neural network so that the model can more accurately reflect the dynamic changes of the power system under small disturbances. The actual operating data of the power system and artificially simulated small disturbance events are considered in the training process to ensure that the model can accurately predict the behavior of the system under different operating conditions.

[0128] In some embodiments, the above-mentioned device also includes a third acquisition unit and a determination unit. The third acquisition unit is used to obtain the similarity between the above-mentioned virtual model and the above-mentioned power system after modeling the above-mentioned virtual model data corresponding to the above-mentioned relevant information to obtain the above-mentioned virtual model; the determination unit is used to determine the accuracy of the above-mentioned intelligent generation model based on the above-mentioned similarity, wherein the above-mentioned similarity and the above-mentioned accuracy are positively correlated.

[0129] In this solution, the interpretability of the trained model can be evaluated. The accuracy of the model can be further evaluated by comparing the similarity between the system state predicted by the neural network model and the actual system state. At the same time, the interpretability of the model can be further improved by analyzing the relationship between the neural network model parameters and the system matrix.

[0130] In some embodiments, the above-mentioned device also includes a fourth acquisition unit and a fourth processing unit, the fourth acquisition unit is used to obtain the initial relevant information of the above-mentioned power system before obtaining the relevant information of the power system, wherein the above-mentioned initial relevant information is the initial operation information of the above-mentioned power equipment; the fourth processing unit is used to preprocess the above-mentioned initial relevant information to obtain the above-mentioned relevant information, wherein the data preprocessing method includes one or more of denoising, standardization and normalization.

[0131] In this scheme, for the initial measured initial relevant information, the collected data is cleaned and preprocessed, including denoising, standardization, normalization and other operations to improve the data quality to ensure that the data quality meets the input requirements of the model.

[0132] The above-mentioned power system modeling device includes a processor and a memory, and the above-mentioned first acquisition unit, first construction unit, first processing unit, second processing unit, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions. The above-mentioned modules are all located in the same processor; or, the above-mentioned modules are located in different processors in the form of any combination.

[0133] The processor includes a kernel, and the kernel retrieves the corresponding program unit from the memory. One or more kernels can be provided, and the problem of poor accuracy of the power system model constructed in the prior art can be solved by adjusting kernel parameters.

[0134] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0135] An embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the power system modeling method.

[0136] Specifically, the power system modeling methods include:

[0137] Step S201, obtaining relevant information of the power system, wherein the relevant information is operation information of the power equipment, wherein the power equipment at least includes one or more of power generation equipment, power transmission equipment, power distribution equipment and power consumption equipment, and the operation information includes one or more of phase angle, speed, transient potential, voltage, current, power and frequency;

[0138] Step S202, constructing an intelligent generation model, wherein the intelligent generation model is obtained by training with a plurality of sets of training data through a neural network algorithm and a numerical integration algorithm, wherein each set of training data in the plurality of sets of training data includes historical related information acquired in a historical time period and historical virtual model data corresponding to the historical related information, wherein the historical virtual model data is virtual model data corresponding to a historical power system;

[0139] Step S203, inputting the above-mentioned relevant information into the above-mentioned intelligent generation model to obtain virtual model data corresponding to the above-mentioned relevant information;

[0140] Step S204, modeling is performed according to the virtual model data corresponding to the relevant information to obtain the virtual model, wherein the virtual model is a virtual model of the power system.

[0141] An embodiment of the present invention provides a processor, and the processor is used to run a program, wherein the modeling method of the power system is executed when the program is run.

[0142] Specifically, the power system modeling methods include:

[0143] Step S201, obtaining relevant information of the power system, wherein the relevant information is operation information of the power equipment, wherein the power equipment at least includes one or more of power generation equipment, power transmission equipment, power distribution equipment and power consumption equipment, and the operation information includes one or more of phase angle, speed, transient potential, voltage, current, power and frequency;

[0144] Step S202, constructing an intelligent generation model, wherein the intelligent generation model is obtained by training with a plurality of sets of training data through a neural network algorithm and a numerical integration algorithm, wherein each set of training data in the plurality of sets of training data includes historical related information acquired in a historical time period and historical virtual model data corresponding to the historical related information, wherein the historical virtual model data is virtual model data corresponding to a historical power system;

[0145] Step S203, inputting the above-mentioned relevant information into the above-mentioned intelligent generation model to obtain virtual model data corresponding to the above-mentioned relevant information;

[0146] Step S204, modeling is performed according to the virtual model data corresponding to the relevant information to obtain the virtual model, wherein the virtual model is a virtual model of the power system.

[0147] An embodiment of the present invention provides a device, comprising a processor, a memory, and a program stored in the memory and executable on the processor, wherein when the processor executes the program, at least the following steps of a method for modeling a power system are implemented:

[0148] Specifically, the power system modeling methods include:

[0149] Step S201, obtaining relevant information of the power system, wherein the relevant information is operation information of the power equipment, wherein the power equipment at least includes one or more of power generation equipment, power transmission equipment, power distribution equipment and power consumption equipment, and the operation information includes one or more of phase angle, speed, transient potential, voltage, current, power and frequency;

[0150] Step S202, constructing an intelligent generation model, wherein the intelligent generation model is obtained by training with a plurality of sets of training data through a neural network algorithm and a numerical integration algorithm, wherein each set of training data in the plurality of sets of training data includes historical related information acquired in a historical time period and historical virtual model data corresponding to the historical related information, wherein the historical virtual model data is virtual model data corresponding to a historical power system;

[0151] Step S203, inputting the above-mentioned relevant information into the above-mentioned intelligent generation model to obtain virtual model data corresponding to the above-mentioned relevant information;

[0152] Step S204, modeling is performed according to the virtual model data corresponding to the relevant information to obtain the virtual model, wherein the virtual model is a virtual model of the power system.

[0153] The devices in this article can be servers, PCs, PADs, mobile phones, etc.

[0154] A computer program product includes a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the power system modeling method in each embodiment of the present application are implemented:

[0155] Specifically, the power system modeling methods include:

[0156] Step S201, obtaining relevant information of the power system, wherein the relevant information is operation information of the power equipment, wherein the power equipment at least includes one or more of power generation equipment, power transmission equipment, power distribution equipment and power consumption equipment, and the operation information includes one or more of phase angle, speed, transient potential, voltage, current, power and frequency;

[0157] Step S202, constructing an intelligent generation model, wherein the intelligent generation model is obtained by training with a plurality of sets of training data through a neural network algorithm and a numerical integration algorithm, wherein each set of training data in the plurality of sets of training data includes historical related information acquired in a historical time period and historical virtual model data corresponding to the historical related information, wherein the historical virtual model data is virtual model data corresponding to a historical power system;

[0158] Step S203, inputting the above-mentioned relevant information into the above-mentioned intelligent generation model to obtain virtual model data corresponding to the above-mentioned relevant information;

[0159] Step S204, modeling is performed according to the virtual model data corresponding to the relevant information to obtain the virtual model, wherein the virtual model is a virtual model of the power system.

[0160] The present application also provides a modeling system, comprising one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing any one of the above-mentioned power system modeling methods.

[0161] Specifically, the power system modeling methods include:

[0162] Step S201, obtaining relevant information of the power system, wherein the relevant information is operation information of the power equipment, wherein the power equipment at least includes one or more of power generation equipment, power transmission equipment, power distribution equipment and power consumption equipment, and the operation information includes one or more of phase angle, speed, transient potential, voltage, current, power and frequency;

[0163] Step S202, constructing an intelligent generation model, wherein the intelligent generation model is obtained by training with a plurality of sets of training data through a neural network algorithm and a numerical integration algorithm, wherein each set of training data in the plurality of sets of training data includes historical related information acquired in a historical time period and historical virtual model data corresponding to the historical related information, wherein the historical virtual model data is virtual model data corresponding to a historical power system;

[0164] Step S203, inputting the above-mentioned relevant information into the above-mentioned intelligent generation model to obtain virtual model data corresponding to the above-mentioned relevant information;

[0165] Step S204, modeling is performed according to the virtual model data corresponding to the relevant information to obtain the virtual model, wherein the virtual model is a virtual model of the power system.

[0166] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0167] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0168] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0169] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0170] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0171] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0172] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0173] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0174] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0175] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0176] 1) The modeling method of the power system of the present application can construct a model through a neural network algorithm and a numerical integration algorithm. The neural network algorithm can learn a large amount of data for rapid learning, while the numerical integration algorithm can accurately simulate various dynamic processes in the power system. Therefore, the intelligent generation model constructed by this scheme has high accuracy, and the accuracy of the obtained power system model is good.

[0177] 2) The modeling device of the power system of the present application can construct a model through a neural network algorithm and a numerical integration algorithm. The neural network algorithm can learn a large amount of data for rapid learning, while the numerical integration algorithm can accurately simulate various dynamic processes in the power system. Therefore, the intelligent generation model constructed by this scheme has a high accuracy, and the accuracy of the obtained power system model is better.

[0178] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for modeling a power system, characterized in that: include: Acquire relevant information of the power system, wherein the relevant information is operation information of the power equipment, wherein the power equipment includes at least one or more of power generation equipment, power transmission equipment, power distribution equipment and power consumption equipment, and the operation information includes one or more of phase angle, rotation speed, transient potential, voltage, current, power and frequency; Constructing an intelligent generation model, wherein the intelligent generation model is obtained by training with a plurality of sets of training data through a neural network algorithm and a numerical integration algorithm, wherein each set of training data in the plurality of sets of training data includes historical related information acquired within a historical time period and historical virtual model data corresponding to the historical related information, wherein the historical virtual model data is virtual model data corresponding to a historical power system; Inputting the relevant information into the intelligent generation model to obtain virtual model data corresponding to the relevant information; Modeling is performed according to the virtual model data corresponding to the relevant information to obtain the virtual model, wherein the virtual model is a virtual model of the power system.

2. The method according to claim 1, characterized in that Before constructing the intelligent generation model, the method further includes: Obtain a differential algebraic equation, wherein the differential algebraic equation is: x represents the state quantity of the power system, represents the derivative of the state quantity, y represents the algebraic quantity, f() represents the differential equation part describing the power system, and g() represents the algebraic equation part describing the power system; The non-state quantities in the differential algebraic equation are removed to obtain a differential equation system, wherein the differential equation system is: The differential equation group is used to construct the intelligent generation model.

3. The method according to claim 2, characterized in that Before constructing the intelligent generation model, the method further includes: According to a first formula, the state quantity of the power system at the target time is calculated, wherein the first formula is: t k represents the target time, x(t k ) represents the state quantity of the power system at the target time, t0 represents the initial time, x(t0) represents the state quantity of the power system at the initial time, f θ (x)=θ1x+θ2, θ1 represents the weight coefficient, and θ2 represents the bias coefficient.

4. The method according to claim 2, characterized in that: Before constructing the intelligent generation model, the method further includes: The sequence expansion formula is used to expand the sample data to obtain the expanded sample data, wherein the sequence expansion formula is: and e,t,i =θ 3,i ·[and o,t-r,i ,and o,t-r+1,i ,...,and o,t+r-1,i ,and o,t+r,i ] T , y e,t,i represents the value of the i-th supplementary quantity measured at time t obtained by expansion, θ 3,i represents the network weight parameter of the i-th quantity measurement, y o,t,i represents the value of the i-th element of the original quantity measurement at time t, t represents the target time, r represents the time window length of each sequence expander, and T represents the matrix transpose symbol.

5. The method according to claim 2, characterized in that: Before constructing the intelligent generation model, the method further includes: Constructing an initial intelligent generation model, wherein the initial intelligent generation model is obtained by training a neural network algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes the historical relevant information obtained within a historical time period and the historical virtual model data corresponding to the historical relevant information; The initial intelligent generation model is optimized by using numerical integration and mean square error loss function to obtain the intelligent generation model.

6. The method according to claim 1, characterized in that After modeling is performed according to the virtual model data corresponding to the relevant information to obtain the virtual model, the method further includes: Acquiring a similarity between the virtual model and the power system; The accuracy of the intelligent generation model is determined according to the similarity, wherein the similarity and the accuracy are positively correlated.

7. The method according to claim 1, characterized in that Before obtaining relevant information of the power system, the method further includes: Acquiring initial relevant information of the power system, wherein the initial relevant information is initial operation information of the power equipment; The initial relevant information is preprocessed to obtain the relevant information, wherein the data preprocessing method includes one or more of denoising, standardization and normalization.

8. A modeling device for an electric power system, characterized in that: include: A first acquisition unit is used to acquire relevant information of the power system, wherein the relevant information is operation information of the power equipment, wherein the power equipment includes at least one or more of power generation equipment, power transmission equipment, power distribution equipment and power consumption equipment, and the operation information includes one or more of phase angle, rotation speed, transient potential, voltage, current, power and frequency; A first construction unit is used to construct an intelligent generation model, wherein the intelligent generation model is obtained by training with a plurality of sets of training data through a neural network algorithm and a numerical integration algorithm, each set of training data in the plurality of sets of training data includes historical relevant information obtained within a historical time period and historical virtual model data corresponding to the historical relevant information, wherein the historical virtual model data is virtual model data corresponding to a historical power system; A first processing unit, configured to input the relevant information into the intelligent generation model to obtain virtual model data corresponding to the relevant information; The second processing unit is used to perform modeling according to the virtual model data corresponding to the relevant information to obtain the virtual model, wherein the virtual model is a virtual model of the power system.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method for modeling a power system according to any one of claims 1 to 7 are implemented.

10. A modeling system, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing the modeling method of the power system described in any one of claims 1 to 7.

Citation Information

Cited By

  • SQL-based equipment operation parameter prediction method, equipment and medium

    CN121412249A

  • Power system high-precision modeling method based on dynamic parameter traceability

    CN121480271A