Data modeling method and system for power system, storage medium and equipment

Abstract

The embodiment of the invention discloses a data modeling method for an electric power system, and the method comprises the steps: obtaining the load data of the electric power system, and the load data of the electric power system comprises power and current; decomposing the power system load data to obtain a plurality of modal functions; matching the power and the current based on the plurality of modal functions; performing power system load multi-model parallel modeling on the plurality of successfully matched powers and currents to obtain a corresponding load model; modeling a plurality of unsuccessfully matched powers and currents by adopting a neural network to obtain a corresponding neural network model; and determining a data comprehensive model of the power system according to the corresponding load model and the corresponding neural network model. According to the method, the accuracy and reliability of power system load modeling are effectively improved by combining modal decomposition and neural network technologies.

Description

Technical Field

[0001] The present invention relates to the technical field of power system load modeling, and in particular to a data modeling method, system, storage medium and equipment for power systems. Background Art

[0002] In power systems, load forecasting and modeling are crucial for power system planning, operation, and control. Accurate load forecasting can help power system operators optimize resource allocation, reduce the risk of power outages, and improve the reliability and efficiency of power systems.

[0003] However, the load of the power system is complex and uncertain, and traditional load modeling methods often have difficulty in accurately capturing the dynamic changes of the load. In particular, in the processing of multi-dimensional data, it is difficult to simultaneously consider multiple factors such as active power, reactive power and current, which limits the accuracy of load forecasting in practical applications.

[0004] Traditional load modeling methods mainly rely on physical models, such as static load models and dynamic load models. These models are usually based on simplified assumptions and cannot accurately describe the complex behavior of the load. In addition, the parameters of these models need to be adjusted according to actual data, and the determination of parameters often requires professional knowledge and experience. Summary of the invention

[0005] Based on this, it is necessary to propose a data modeling method for power systems to address the above problems.

[0006] A data modeling method for an electric power system, the method comprising:

[0007] Acquiring power system load data, wherein the power system load data includes power and current;

[0008] Decomposing the power system load data to obtain a plurality of modal functions;

[0009] Matching power and current based on the plurality of modal functions;

[0010] For several successfully matched powers and currents, multiple parallel models of power system load are modeled to obtain corresponding load models; for several unmatched powers and currents, neural network models are used to model to obtain corresponding neural network models;

[0011] A data integration model of the power system is determined according to the corresponding load model and the corresponding neural network model.

[0012] In the above scheme, the power system load data is obtained, and the power system load data includes power and current, and specifically includes:

[0013] Collecting power system load data in real time, wherein the power system load data includes active power, reactive power, voltage and current;

[0014] The power system load data is cleaned and normalized.

[0015] In the above solution, the matching of power and current based on the plurality of modal functions specifically includes:

[0016] Obtain the ideal matching value of the current component in the modal function;

[0017] determining a root mean square error between a current component in each modal function and the ideal matching value;

[0018] The matching results of power and current are determined based on the modal function with the smallest RMS error.

[0019] In the above solution, the step of obtaining the ideal matching value of the current component in the modal function specifically includes:

[0020] The modal function is processed so that it satisfies:

[0021] P=P1+P2+...+P n ,

[0022] Q=Q1+Q2+...+Q n ,

[0023] I=I1+I2+...+I n

[0024] Among them, P represents active power, Q represents reactive power, I represents current, and P i Represents the active power component, Q i Represents the reactive power component, I i represents the current component;

[0025] According to the active power and reactive power, the ideal matching value of the current component in the modal function is obtained.

[0026]

[0027] Among them, I p Represents active current, I Q Indicates reactive current.

[0028] In the above solution, determining the data integration model of the power system according to the corresponding load model and the corresponding neural network model specifically includes:

[0029] F(x)=α1f1(x)+α2f2(x)+...+α n fn (x)+βS(x)

[0030] Among them, F(x) represents the data synthesis model, f i (x) represents the load model, and S(x) represents the neural network model.

[0031] The above scheme also includes processing the data comprehensive model of the power system to obtain the relationship between active power, reactive power, voltage and current:

[0032] make

[0033] The data synthesis model of the power system is transformed into:

[0034]

[0035] Among them, β, α i Represents the weight coefficient of the neural network model; P Mi (i=1,2,...,n) represents the active power component of the i-th model after successful matching; Q Mi (i=1,2,...,n) represents the reactive power component of the i-th model after successful matching; U Mi (i=1,2,...,n) represents the voltage of the i-th model after successful matching. U M1 =...=U Mn ;I Mi (i=1,2,...,n) represents the current component of the i-th model after successful matching; P(U,I) and Q(U,I) represent the comprehensive load active power and reactive power models.

[0036] The above scheme also includes optimizing the weight coefficients of the neural network model:

[0037] Define the central frequency values ​​of the neural network model as w1,w2,...,w n ;

[0038] The initial value of the weight coefficient β is set to 1, and the weight coefficient α is determined according to the following formula: i :

[0039]

[0040] The error of the neural network model is determined according to the following formula;

[0041]

[0042] Where n represents the number of samples; is the active and reactive power calculated at time i, Pm,i , Q m,i are the active and reactive powers measured at time i;

[0043] When the error of the neural network model meets the threshold range, the current weight coefficient is output.

[0044] The present application also proposes a data modeling system for a power system, the system comprising: a load data acquisition unit, a data processing unit, a matching unit and a modeling unit;

[0045] The load data acquisition unit is used to acquire power system load data, wherein the power system load data includes power and current;

[0046] The data processing unit is used to decompose the power system load data to obtain a plurality of modal functions;

[0047] The matching unit is used to match power and current based on the plurality of modal functions;

[0048] The modeling unit is used to perform parallel modeling of power system load multi-models for several successfully matched powers and currents to obtain corresponding load models; use neural networks to model several unsuccessfully matched powers and currents to obtain corresponding neural network models; and determine the data comprehensive model of the power system based on the corresponding load models and the corresponding neural network models.

[0049] The present application also proposes a readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the following steps:

[0050] Acquiring power system load data, wherein the power system load data includes power and current;

[0051] Decomposing the power system load data to obtain a plurality of modal functions;

[0052] Matching power and current based on the plurality of modal functions;

[0053] For several successfully matched powers and currents, multiple parallel models of power system load are modeled to obtain corresponding load models; for several unmatched powers and currents, neural network models are used to model to obtain corresponding neural network models;

[0054] A data integration model of the power system is determined according to the corresponding load model and the corresponding neural network model.

[0055] The present application also proposes a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the following steps of the computer program:

[0056] Acquiring power system load data, wherein the power system load data includes power and current;

[0057] Decomposing the power system load data to obtain a plurality of modal functions;

[0058] Matching power and current based on the plurality of modal functions;

[0059] For several successfully matched powers and currents, multiple parallel models of power system load are modeled to obtain corresponding load models; for several unmatched powers and currents, neural network models are used to model to obtain corresponding neural network models;

[0060] A data integration model of the power system is determined according to the corresponding load model and the corresponding neural network model.

[0061] The use of the embodiments of the present invention has the following beneficial effects: firstly, the load data of the power system is obtained, and the load data of the power system includes power and current; the load data of the power system is decomposed to obtain several modal functions; the power and current are matched based on the several modal functions; the power system load multi-model parallel modeling is performed for several successfully matched powers and currents to obtain the corresponding load model; the power and current that are not successfully matched are modeled using a neural network to obtain the corresponding neural network model; the data comprehensive model of the power system is determined according to the corresponding load model and the corresponding neural network model. The present invention decomposes complex load signals into multiple characteristic modes, identifies load components, matches power and current data, establishes corresponding load models, performs neural network modeling on unmatched data, handles nonlinear relationships, and finally integrates the load model and the neural network model to construct a comprehensive model of power system data. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0063] in:

[0064] Figure 1 A schematic diagram of a flow chart of a data modeling method for a power system in one embodiment;

[0065] Figure 2 A schematic diagram of parallel modeling of multiple models of power system loads in one embodiment. DETAILED DESCRIPTION

[0066] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0067] In the following description, a large number of specific details are given in order to provide a more thorough understanding of the present invention; however, it is obvious to those skilled in the art that the present invention can be implemented without one or more of these details; in other examples, in order to avoid confusion with the present invention, some technical features known in the art are not described, and it should be understood that the present invention can be implemented in different forms and should not be interpreted as limited to the embodiments presented here; on the contrary, providing these embodiments will make the disclosure thorough and complete and fully convey the scope of the present invention to those skilled in the art.

[0068] The purpose of the terms used herein is only to describe specific embodiments and is not intended to be limiting of the present invention. When used herein, the singular forms "a", "an" and "said / the" are also intended to include the plural forms, unless the context clearly indicates otherwise, and it should also be understood that the terms "consisting of" and / or "comprising", when used in this specification, determine the presence of the features, integers, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, parts and / or groups. When used herein, the term "and / or" includes any and all combinations of the relevant listed items.

[0069] In order to thoroughly understand the present invention, a detailed structure will be presented in the following description to illustrate the technical solution proposed by the present invention; optional embodiments of the present invention are described in detail as follows, but in addition to these detailed descriptions, the present invention may also have other implementation methods.

[0070] like Figure 1 As shown, in one embodiment, a data modeling method for a power system is provided, and the data modeling method for a power system includes steps S101 to S106, which are described in detail as follows:

[0071] 101. Obtain power system load data, where the power system load data includes power and current;

[0072] This step collects key data on the operation of the power system, including active power, reactive power and current, providing a basis for subsequent modeling and analysis.

[0073] In some embodiments, the power system load data is obtained, the power system load data includes power and current, and specifically includes:

[0074] Real-time collection of power system load data, including active power, reactive power, voltage and current;

[0075] Perform data cleaning and normalization on power system load data.

[0076] 102. Decompose the load data of the power system to obtain several modal functions;

[0077] By decomposing load data, complex load signals can be decomposed into multiple modal functions with different characteristics, which helps to identify different components of the load, such as fundamental load, harmonic load, and random fluctuations.

[0078] In some embodiments, the power system load data is decomposed using an empirical mode decomposition algorithm:

[0079] a) Find the maximum and minimum points of the given time series function {x(t)} (modal function);

[0080] b) Connect each maximum point with a curve through cubic spline interpolation to obtain the upper envelope x max (t), each minimum point is connected to obtain the lower envelope x min (t);

[0081] c) Calculate the mean of the corresponding upper and lower envelopes and record it as m1:

[0082]

[0083] d) Subtract the mean to get the new function:

[0084] h1(t)=x(t)-m1 (2)

[0085] e) Replace the original x(t) with the new function h1(t) and repeat the above steps to get h k (t)(k=2,3,...) stops when it meets the IMF condition, and defines:

[0086] IMF1(t)=h k (t) (3)

[0087] f) Separate IMF1 from x(t) to obtain r1.

[0088] r1=x(t)-IMF1 (4)

[0089] Repeat n times to obtain n IMF components and residual component r of x(t) n for:

[0090] r n =r n-1 -IMF n (5)

[0091] When n is a monotone function. EMD decomposition stops and the following result is obtained:

[0092]

[0093] It can be seen that through steps af, the modal function {x(t)} is decomposed into mutually orthogonal eigenmode functions (modal functions) and a trend term.

[0094] 103. Matching power and current based on several modal functions;

[0095] By matching the modal functions, the power and current data can be correlated to more accurately identify the composition and characteristics of the load.

[0096] In some embodiments, power and current are matched based on several modal functions, including:

[0097] Obtain the ideal matching value of the current component in the modal function;

[0098] Determine the root mean square error between the current component in each modal function and the ideal matching value;

[0099] The matching results of power and current are determined based on the modal function with the smallest RMS error.

[0100] By calculating the root mean square error, the difference between each modal function and the ideal value can be quantified, so that the modal function with the smallest error can be selected as the matching result. The modal function with the smallest error usually represents the main component of the load, such as the fundamental load or the harmonic load. Finally, through matching, the complex load signal can be decomposed into simpler components, thereby simplifying the complexity of the model.

[0101] In some embodiments, obtaining an ideal matching value of a current component in a modal function specifically includes:

[0102] The modal function is processed so that it satisfies:

[0103] P=P1+P2+...+P n ,

[0104] Q=Q1+Q2+...+Q n ,

[0105] I=I1+I2+...+I n

[0106] Among them, P represents active power, Q represents reactive power, I represents current, and P i Represents the active power component, Q i Represents the reactive power component, I i represents the current component;

[0107] Obtain the ideal matching value of the current component in the modal function based on active power and reactive power

[0108]

[0109] Among them, I p Represents active current, I Q Indicates reactive current.

[0110] Preferably, the ideal matching value of the current is obtained by using the active current and the reactive current. The ideal matching value of the current is combined with each decomposed current I j Compare and calculate the root mean square error RMSD of each matching result j :

[0111]

[0112] in is the ideal current matching value corresponding to the i-th current component and the n-th sampling point, I j,n is the actual current value corresponding to the jth current component and the nth sampling point;

[0113] After calculating the root mean square error of each matching result, take the minimum RMSD to achieve the matching effect; if The corresponding min(RMSD) = RMSD j , which means P i , Q i with I j Has the highest matching degree.

[0114] 104. Perform parallel modeling of multiple load models of the power system for the successfully matched powers and currents to obtain the corresponding load model;

[0115] like Figure 2The figure shows a schematic diagram of multi-model parallel modeling of power system load. The voltage of the multi-model parallel model remains unchanged. The curves of load active power P, reactive power Q, current I, etc. can be decomposed and then modeled separately to obtain a load model. The data before and after decomposition satisfy P = P1 + P2 + ... + P n , Q=Q1+Q2+...+Q n , I=I1+I2+...+I n .

[0116] 105. Modeling a number of powers and currents that have not been successfully matched using a neural network to obtain a corresponding neural network model;

[0117] For power and current data that cannot be matched, neural networks can be used for modeling. Neural networks can learn complex nonlinear relationships to more accurately describe the dynamic changes and uncertainties of the load.

[0118] 106. Determine the data integration model of the power system according to the corresponding load model and the corresponding neural network model.

[0119] By integrating the load model and the neural network model, a comprehensive power system data model can be established, which can be used for applications such as load forecasting, system planning and control.

[0120] In some embodiments, determining a data integration model of a power system according to a corresponding load model and a corresponding neural network model specifically includes:

[0121] F(x)=α1f1(x)+α2f2(x)+...+α n f n (x)+βS(x)

[0122] Among them, F(x) represents the data synthesis model, f i (x) represents the load model, and S(x) represents the neural network model.

[0123] In some embodiments, the method further includes processing a data integration model of the power system to obtain the relationship between active power, reactive power, voltage and current:

[0124] make

[0125] Transform the data synthesis model of the power system into:

[0126]

[0127] Among them, β, α i Represents the weight coefficient of the neural network model; P Mi(i=1,2,...,n) represents the active power component of the i-th model after successful matching; Q Mi (i=1,2,...,n) represents the reactive power component of the i-th model after successful matching; U Mi (i=1,2,...,n) represents the voltage of the i-th model after successful matching. U M1 =...=U Mn ;I Mi (i=1,2,...,n) represents the current component of the i-th model after successful matching; P(U,I) and Q(U,I) represent the comprehensive load active power and reactive power models.

[0128] In some embodiments, the method further includes optimizing the weight coefficients of the neural network model:

[0129] Define the central frequency values ​​of the neural network model as w1,w2,...,w n ;

[0130] The initial value of the weight coefficient β is set to 1, and the weight coefficient α is determined according to the following formula: i :

[0131]

[0132] The error of the neural network model is determined according to the following formula;

[0133]

[0134] Where n represents the number of samples; represents the active and reactive power calculated at time i, P m,i , Q m,i represents the active and reactive power measured at time i;

[0135] When the error of the neural network model meets the threshold range, the current weight coefficient is output.

[0136] In summary, this application forms a comprehensive model by integrating the load model and the neural network model, and uses a test data set to evaluate the performance of the comprehensive model, including accuracy, robustness, etc. According to the evaluation results, the comprehensive model is optimized to improve the prediction ability and adaptability of the model, which can effectively capture the multi-scale characteristics of the load, making power load modeling more flexible, improving the accuracy and robustness of load modeling, and making it widely used in the fields of load forecasting, power generation scheduling and power grid control of power systems.

[0137] The present application also proposes a data modeling system for a power system, the system comprising: a load data acquisition unit, a data processing unit, a matching unit and a modeling unit;

[0138] A load data acquisition unit, used to acquire power system load data, the power system load data including power and current;

[0139] A data processing unit, used for decomposing the load data of the power system and obtaining a number of modal functions;

[0140] A matching unit for matching power and current based on a number of modal functions;

[0141] The modeling unit is used to perform parallel modeling of power system load multi-models for several successfully matched powers and currents to obtain corresponding load models; to use neural networks to model several unsuccessfully matched powers and currents to obtain corresponding neural network models; and to determine the data comprehensive model of the power system according to the corresponding load model and the corresponding neural network model.

[0142] The present application also proposes a readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the following steps:

[0143] Obtaining power system load data, which includes power and current;

[0144] Decompose the power system load data to obtain several modal functions;

[0145] Matching power and current based on several modal functions;

[0146] For several successfully matched powers and currents, multiple parallel models of power system load are modeled to obtain corresponding load models; for several unmatched powers and currents, neural network models are used to model to obtain corresponding neural network models;

[0147] The data integration model of the power system is determined according to the corresponding load model and the corresponding neural network model.

[0148] The present application also proposes a computer device, including a memory and a processor, wherein the memory stores a computer program, and the computer program is executed by the processor to perform the following steps:

[0149] Obtaining power system load data, which includes power and current;

[0150] Decompose the power system load data to obtain several modal functions;

[0151] Matching power and current based on several modal functions;

[0152] For several successfully matched powers and currents, multiple parallel models of power system load are modeled to obtain corresponding load models; for several unmatched powers and currents, neural network models are used to model to obtain corresponding neural network models;

[0153] The data integration model of the power system is determined according to the corresponding load model and the corresponding neural network model.

[0154] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0155] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0156] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the present application. It should be pointed out that, for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. The above disclosure is only a preferred embodiment of the present invention, and of course it cannot be used to limit the scope of the rights of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope covered by the present invention.

Claims

1. A data modeling method for a power system, characterized in that: The method comprises: Acquiring power system load data, wherein the power system load data includes power and current; Decomposing the power system load data to obtain a plurality of modal functions; Matching power and current based on the plurality of modal functions; For several successfully matched powers and currents, multiple parallel models of power system load are modeled to obtain the corresponding load model; A neural network is used to model several powers and currents that have not been successfully matched, and a corresponding neural network model is obtained; A data integration model of the power system is determined according to the corresponding load model and the corresponding neural network model.

2. The data modeling method for power system according to claim 1, characterized in that: The acquiring of power system load data, wherein the power system load data includes power and current, specifically includes: Collecting power system load data in real time, wherein the power system load data includes active power, reactive power, voltage and current; The power system load data is cleaned and normalized.

3. The data modeling method for power system according to claim 1, characterized in that: The matching of power and current based on the plurality of modal functions specifically includes: Obtain the ideal matching value of the current component in the modal function; determining a root mean square error between a current component in each modal function and the ideal matching value; The matching results of power and current are determined based on the modal function with the smallest RMS error.

4. The data modeling method for power system according to claim 3, characterized in that: The obtaining of the ideal matching value of the current component in the modal function specifically includes: The modal function is processed so that it satisfies: P=P1+P2+...+P n , Q=Q1+Q2+...+Q n , I=I1+I2+...+I n Among them, P represents active power, Q represents reactive power, I represents current, and P i Represents the active power component, Q i Represents the reactive power component, I i represents the current component; According to the active power and reactive power, the ideal matching value of the current component in the modal function is obtained. Among them, I p Represents active current, I Q Indicates reactive current.

5. The data modeling method for power system according to claim 1, characterized in that: Determining the data integration model of the power system according to the corresponding load model and the corresponding neural network model specifically includes: F(x)=α1f1(x)+α2f2(x)+...+α n f n (x)+βS(x) Among them, F(x) represents the data synthesis model, f i (x) represents the load model, and S(x) represents the neural network model.

6. The data modeling method for power system according to claim 5, characterized in that: The method also includes processing the data synthesis model of the power system to obtain the relationship between active power, reactive power, voltage and current: make The data synthesis model of the power system is transformed into: Among them, β, α i Represents the weight coefficient of the neural network model; P Mi (i=1,2,...,n) represents the active power component of the i-th model after successful matching; Q Mi (i=1,2,...,n) represents the reactive power component of the i-th model after successful matching; U Mi (i=1,2,...,n) represents the voltage of the i-th model after successful matching. U M1 =...=U Mn ;I Mi (i=1,2,...,n) represents the current component of the i-th model after successful matching; P(U,I) and Q(U,I) represent the comprehensive load active power and reactive power model.

7. The data modeling method for power system according to claim 6, characterized in that: It also includes optimizing the weight coefficients of the neural network model: Define the central frequency values ​​of the neural network model as w1,w2,...,w n ; The initial value of the weight coefficient β is set to 1, and the weight coefficient α is determined according to the following formula: i : The error of the neural network model is determined according to the following formula; Where n represents the number of samples; represents the active and reactive power calculated at time i, P m,i , Q m,i represents the active and reactive power measured at time i; When the error of the neural network model meets the threshold range, the current weight coefficient is output.

8. A data modeling system for a power system, characterized in that: The system comprises: a load data acquisition unit, a data processing unit, a matching unit and a modeling unit; The load data acquisition unit is used to acquire power system load data, wherein the power system load data includes power and current; The data processing unit is used to decompose the power system load data to obtain a plurality of modal functions; The matching unit is used to match power and current based on the plurality of modal functions; The modeling unit is used to perform parallel modeling of power system load multi-models for several successfully matched powers and currents to obtain corresponding load models; use neural networks to model several unsuccessfully matched powers and currents to obtain corresponding neural network models; and determine the data comprehensive model of the power system based on the corresponding load models and the corresponding neural network models.

9. A readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor is caused to execute the steps of the method according to any one of claims 1 to 7.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method according to any one of claims 1 to 7.

Images