Load prediction method and system, electronic equipment, storage medium and program product

Abstract

The invention provides a load prediction method and system, electronic equipment, a storage medium and a program product. The method comprises the steps of obtaining a multi-source circuit load influence parameter and storing the parameter in a database; performing hypothesis period test on the circuit load influence parameter, performing frequency domain decomposition on the circuit load influence parameter and a total load period, and storing a matching result into a parameter library; fitting the database and the parameter library to obtain a plurality of pre-selection models, calculating expected risks of the pre-selection models, and performing VC dimension analysis on the pre-selection models to obtain confidence intervals of the pre-selection models; screening out the pre-selection models of which the expected risks are located outside the confidence interval, performing principal component analysis on the remaining pre-selection models to obtain a basis kernel function, and planning the minimum value of a prediction error by taking the value domain of the basis kernel function as a limiting condition; and selecting the residual pre-selected model with the minimum prediction error as a target model, and calculating the actual error volatility of the target model to adjust the prediction result of the target model. According to the invention, the accuracy of load prediction is improved, and the universality, flexibility and timeliness are improved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of power systems, and particularly to a load forecasting method, system, electronic device, storage medium, and program product. Background Art

[0002] This section aims to provide background or context for the embodiments of the present disclosure stated in the claims. The description herein is not admitted to be prior art merely by including it in this section.

[0003] The power system is an important part of energy application. In the power system, the process of estimating the electric load that will be demanded in the power system or energy system within a certain period in the future based on historical data, environmental factors, and other relevant information is called load forecasting. Through load forecasting, power generation units can reasonably arrange power generation plans and power dispatching to meet the actual demand.

[0004] There are many reasons affecting the load of the power system. Even the same reason has different manifestations in different power consumption places, making it very difficult to predict the load in complex scenarios. The load forecasting models in the related technologies predict the future power consumption load by analyzing factors such as historical load data, weather conditions, and holidays. However, the influencing factors are not considered comprehensively enough, and the specific influence degree of each factor on the load cannot be accurately judged, resulting in inaccurate load forecasting models and poor model versatility.

[0005] In addition, the source function selected by the load model determines the prediction effect of the model to a certain extent. The selection of the source function depends on personal experience and local historical power consumption characteristics and cannot be quantitatively adjusted according to real-time data, affecting the flexibility and timeliness of model prediction and causing interference to the load forecasting results. Summary of the Invention

[0006] In view of this, an object of the present disclosure is to provide a load forecasting method, system, electronic device, storage medium, and program product, which can at least solve one of the technical problems in the related technologies to a certain extent.

[0007] Based on the above object, the first aspect of the exemplary embodiment of the present disclosure provides a load forecasting method, including:

[0008] Obtain circuit load impact parameters from the power system and the operating environment of power equipment, and store the circuit load impact parameters in a database;

[0009] Perform a hypothesis periodic test on the circuit load impact parameters to obtain the test frequency of the circuit load impact parameters, obtain the variation function of the total circuit load over time, decompose the variation function based on the test frequency to obtain the component functions corresponding to the circuit load impact parameters, calculate the integral of the component functions within the assumed period, and store the circuit load impact parameters with integrals higher than the threshold in the parameter library;

[0010] Fit the database and the parameter library based on several methods to obtain several candidate models, calculate the expected risk of the candidate models according to the probability distribution of the simulation results of the candidate models, perform VC dimension analysis on the candidate models to obtain the confidence interval of the candidate models;

[0011] Screen out the candidate models whose expected risks are outside the confidence interval, perform principal component analysis on the remaining candidate models, use the function of the parameters in the parameter library corresponding to the principal components as the kernel function of the remaining candidate models, and perform linear programming on the minimum value of the prediction error of the remaining candidate models with the value range of the kernel function as the constraint condition;

[0012] Select the remaining candidate model with the smallest prediction error as the target model, calculate the actual error volatility of the target model, and adjust the prediction result of the target model based on the actual error volatility.

[0013] In some exemplary embodiments, the obtaining the circuit load impact parameters from the power system and the operating environment of the power equipment and storing the circuit load impact parameters in the database includes:

[0014] Obtain electrical impact parameters from the power system and environmental impact parameters from the operating environment of the power equipment, where the circuit load impact parameters include the electrical impact parameters and the environmental impact parameters;

[0015] Serialize the electrical impact parameters and the environmental impact parameters in chronological order to obtain the time series corresponding to the electrical impact parameters and the environmental impact parameters, and store the time series in the database.

[0016] In some exemplary embodiments, the performing a hypothesis periodic test on the circuit load impact parameters to obtain the test frequency of the circuit load impact parameters, obtaining the variation function of the total circuit load over time, decomposing the variation function based on the test frequency to obtain the component functions corresponding to the circuit load impact parameters, calculating the integral of the component functions within the assumed period, and storing the circuit load impact parameters with integrals higher than the threshold in the parameter library includes:

[0017] Calculate the mean and variance of each element in the time series;

[0018] Perform a hypothesis periodic test on the time series based on the mean value and the variance to obtain the test period of the time series;

[0019] Based on the test period of the time series, obtain the test frequency of the circuit load influence parameter;

[0020] Obtain the variation function of the total circuit load over time;

[0021] Decompose the variation function based on the test frequency to obtain the component function corresponding to the circuit load influence parameter;

[0022] Calculate the integral of the component function within the assumed period, and store the circuit load influence parameters with integrals higher than the threshold in the parameter library.

[0023] In some exemplary embodiments, the method of fitting the database and the parameter library in several ways to obtain several preselected models, calculating the expected risk of the preselected models according to the probability distribution of the simulation results of the preselected models, and performing VC dimension analysis on the preselected models to obtain the confidence interval of the preselected models includes:

[0024] Taking the data of the database as the output quantity and the parameters of the parameter library as the input quantity, fitting the database and the parameter library in several ways to obtain several preselected models;

[0025] Calculate the simulation results of the preselected models, and calculate the expected risk of the preselected models according to the probability distribution of the simulation results of the preselected models;

[0026] Calculate the VC dimension of the preselected models to obtain the confidence interval of the preselected models.

[0027] In some exemplary embodiments, the method of screening out the preselected models with the expected risk outside the confidence interval, performing principal component analysis on the remaining preselected models, using the function of the parameters of the parameter library corresponding to the principal components as the kernel function of the remaining preselected models, and performing linear programming on the minimum value of the prediction error of the remaining preselected models with the value range of the kernel function as the constraint condition includes:

[0028] Screen out the preselected models with the expected risk outside the confidence interval, perform principal component analysis on the remaining preselected models, and use the main function of the function set selected when fitting the remaining preselected models as the kernel function of the remaining preselected models;

[0029] Taking the value range of the kernel function as the constraint condition, perform linear programming on the minimum value of the prediction error of the remaining preselected models, and record the minimum value as the optimization parameter of the remaining preselected models.

[0030] In some exemplary embodiments, selecting the remaining preselected model with the smallest prediction error as the target model, calculating the actual error volatility of the target model, and adjusting the prediction result of the target model based on the actual error volatility includes:

[0031] Selecting the remaining preselected model with the smallest prediction error as the target model, and using the target model to perform load forecasting for the power system in the next cycle;

[0032] During the load forecasting process in the next cycle, detecting the error between the prediction result and the measured result, generating an error sampling sequence, performing stability analysis on the error sampling sequence, and obtaining the actual error volatility of the target model;

[0033] Predicting the next element value of the error sampling sequence based on the actual error volatility, and adjusting the prediction result of the target model based on the next element value;

[0034] The method further includes:

[0035] When the actual error volatility deviates from the preset range, ending the current cycle and replacing the target model in the next cycle.

[0036] Based on the same inventive concept, a second aspect of the exemplary embodiments of the present disclosure provides a load forecasting system, including:

[0037] A factor sensing module, configured to obtain circuit load impact parameters from the power system and the operating environment of power equipment, and store the circuit load impact parameters in a database;

[0038] A relevant screening module, configured to perform a hypothesis period test on the circuit load impact parameters, obtain the test frequency of the circuit load impact parameters, obtain the variation function of the total circuit load over time, decompose the variation function based on the test frequency, obtain the component function corresponding to the circuit load impact parameters, calculate the integral of the component function within the hypothesis period, and store the circuit load impact parameters with an integral higher than the threshold in a parameter library;

[0039] A function optimization module, configured to fit the database and the parameter library based on several methods to obtain several preselected models, calculate the expected risk of the preselected models according to the probability distribution of the simulation results of the preselected models, perform VC dimension analysis on the preselected models, and obtain the confidence interval of the preselected models;

[0040] The target regression module is configured to screen out the preselected models for which the expected risk lies outside the confidence interval, perform principal component analysis on the remaining preselected models, use a function of the parameters in the parameter library corresponding to the principal components as the base kernel function of the remaining preselected models, and perform linear programming on the minimum value of the prediction error of the remaining preselected models with the value range of the base kernel function as the constraint condition;

[0041] The error correction module is configured to select the remaining preselected model with the smallest prediction error as the target model, calculate the actual error volatility of the target model, and adjust the prediction result of the target model based on the actual error volatility.

[0042] Based on the same inventive concept, a third aspect of the exemplary embodiments of the present disclosure provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the method described in the first aspect is implemented.

[0043] Based on the same inventive concept, a fourth aspect of the exemplary embodiments of the present disclosure provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the method described in the first aspect.

[0044] Based on the same inventive concept, a fifth aspect of the exemplary embodiments of the present disclosure provides a computer program product including computer program instructions that, when run on a computer, cause the computer to execute the method described in the first aspect.

[0045] As can be seen from the above, the load prediction method, system, electronic device, storage medium, and program product provided by the embodiments of the present disclosure, the method includes: obtaining circuit load impact parameters from the power system and the operating environment of power equipment, and storing the circuit load impact parameters in a database; performing a hypothesis period test on the circuit load impact parameters to obtain the test frequency of the circuit load impact parameters, obtaining the variation function of the total circuit load over time, decomposing the variation function based on the test frequency to obtain the component functions corresponding to the circuit load impact parameters, calculating the integral of the component functions within the hypothesis period, and storing the circuit load impact parameters with integrals higher than the threshold in a parameter library; fitting the database and the parameter library in several ways to obtain several candidate models, calculating the expected risk of the candidate models according to the probability distribution of the simulation results of the candidate models, performing VC dimension analysis on the candidate models to obtain the confidence interval of the candidate models; screening out the candidate models with expected risks outside the confidence interval, performing principal component analysis on the remaining candidate models, using the function of the parameters in the parameter library corresponding to the principal components as the kernel function of the remaining candidate models, and performing linear programming on the minimum value of the prediction error of the remaining candidate models with the value range of the kernel function as the constraint condition; selecting the remaining candidate model with the smallest prediction error as the target model, calculating the actual error volatility of the target model, and adjusting the prediction result of the target model based on the actual error volatility. The present disclosure improves the accuracy of load prediction, and also improves the versatility, flexibility, and timeliness. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the present disclosure or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the drawings in the following description are only the embodiments of the present disclosure. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0047] Figure 1 It is a schematic flowchart of a load prediction method provided by an exemplary embodiment of the present disclosure;

[0048] Figure 2 It is a schematic structural diagram of a load prediction system provided by an exemplary embodiment of the present disclosure;

[0049] Figure 3 It is a schematic structural diagram of an electronic device provided by an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] It is understandable that before using the technical solutions disclosed in the embodiments of the present application, the types, usage scopes, usage scenarios, etc. of the personal information involved in the present application should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

[0051] For example, when responding to receiving an active request from a user, a prompt message is sent to the user to clearly prompt the user that the operation requested by the user will require obtaining and using the user's personal information. Thus, the user can autonomously choose whether to provide personal information to software or hardware such as an electronic device, an application program, a server, or a storage medium that performs the operations of the technical solution of the present application according to the prompt message.

[0052] As an optional but non-limiting implementation manner, the manner of sending a prompt message to the user in response to receiving an active request from the user may be, for example, in the form of a pop-up window, and the prompt message may be presented in text in the pop-up window. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0053] It is understandable that the above process of notifying and obtaining the user's authorization is only illustrative and does not limit the implementation manner of the present application. Other manners that meet relevant laws and regulations can also be applied to the implementation manner of the present application.

[0054] It is understandable that the data involved in the technical solution (including but not limited to the data itself, the acquisition or use of the data) should comply with the requirements of the corresponding laws, regulations and related provisions.

[0055] To make the purpose, technical solution and advantages of the present disclosure clearer and more understandable, the principles and spirit of the present disclosure will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are given only to enable those skilled in the art to better understand and then implement the present disclosure, and do not limit the scope of the present disclosure in any way. On the contrary, these embodiments are provided to make the present disclosure more thorough and complete, and to be able to convey the scope of the present disclosure completely to those skilled in the art.

[0056] In this article, it should be understood that any number of elements in the drawings is for illustration rather than limitation, and any naming is only for distinction and does not have any limiting meaning.

[0057] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present disclosure should have the ordinary meanings understood by those of ordinary skill in the field to which the present disclosure belongs. The terms "first", "second" and similar words used in the embodiments of the present disclosure do not denote any order, quantity or importance, but are only used to distinguish different components. Words such as "including" or "comprising" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. Words such as "connected" or "linked" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Upper", "lower", "left", "right", etc. are only used to represent relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly. The article "a" or "an" before an element does not exclude the existence of multiple such elements.

[0058] Next, with reference to several representative embodiments of the present disclosure, the principles and spirits of the present disclosure will be elaborated in detail.

[0059] As described in the background art, the power system is an important part of energy application. In the power system, the process of estimating the power load that will be demanded in the power system or energy system within a certain period of time in the future based on historical data, environmental factors, and other relevant information is called load forecasting. Through load forecasting, power generation units can reasonably arrange power generation plans and power dispatching to meet actual demands.

[0060] There are many reasons that affect the load of the power system. Even the same reason has different manifestations in different power consumption places, making load forecasting in complex scenarios very difficult.

[0061] The load forecasting models in the related art predict the future power consumption load by analyzing factors such as historical load data, weather conditions, and holidays.

[0062] However, the inventors of the present disclosure have found that the influencing factors considered in the load forecasting models in the related art are not comprehensive enough, and the specific influence degrees of various factors on the load cannot be accurately judged, resulting in inaccurate load forecasting models and poor model versatility.

[0063] In addition, the inventors of the present disclosure have also found that the source function selected by the load model determines the prediction effect of the model to a certain extent, and the selection of the source function depends on personal experience and local historical power consumption characteristics, and cannot be quantitatively adjusted according to real-time data, affecting the flexibility and timeliness of model prediction and causing interference to the load forecasting results.

[0064] To solve the above problems, the present disclosure provides a load forecasting scheme, which specifically includes: obtaining circuit load impact parameters from the power system and the operating environment of power equipment, and storing the circuit load impact parameters in a database; performing a hypothesis period test on the circuit load impact parameters to obtain the test frequency of the circuit load impact parameters, obtaining the variation function of the total circuit load over time, decomposing the variation function based on the test frequency to obtain the component functions corresponding to the circuit load impact parameters, calculating the integral of the component functions within the hypothesis period, and storing the circuit load impact parameters with integrals higher than the threshold in a parameter library; fitting the database and the parameter library in several ways to obtain several candidate models, calculating the expected risk of the candidate models according to the probability distribution of the simulation results of the candidate models, performing a VC dimension analysis on the candidate models to obtain the confidence interval of the candidate models; screening out the candidate models whose expected risks are outside the confidence interval, performing a principal component analysis on the remaining candidate models, using the function of the parameters in the parameter library corresponding to the principal components as the kernel function of the remaining candidate models, and performing a linear programming on the minimum value of the prediction error of the remaining candidate models with the value range of the kernel function as the constraint condition; selecting the remaining candidate model with the smallest prediction error as the target model, calculating the actual error volatility of the target model, and adjusting the prediction result of the target model based on the actual error volatility. The present disclosure improves the accuracy of load forecasting, and also improves the versatility, flexibility and timeliness.

[0065] After introducing the basic principle of the present disclosure, the various non-limiting embodiments of the present disclosure will be specifically introduced below.

[0066] Reference Figure 1 , which is a schematic flowchart of a load forecasting method provided by an exemplary embodiment of the present disclosure.

[0067] The load forecasting method includes the following steps:

[0068] Step S110, obtaining circuit load impact parameters from the power system and the operating environment of power equipment, and storing the circuit load impact parameters in a database.

[0069] Specifically, sensors are set in the power system and the operating environment of power equipment, parameters affecting the circuit load are obtained from multiple channels, the collected parameters are cleaned, and each impact parameter is stored in the database in chronological order.

[0070] In some exemplary embodiments, the obtaining circuit load impact parameters from the power system and the operating environment of power equipment, and storing the circuit load impact parameters in a database includes:

[0071] Obtain electrical influence parameters from the power system and environmental influence parameters from the operating environment of the power equipment, where the circuit load influence parameters include the electrical influence parameters and the environmental influence parameters;

[0072] Serialize the electrical influence parameters and the environmental influence parameters in chronological order to obtain the time series corresponding to the electrical influence parameters and the environmental influence parameters, and store the time series in a database.

[0073] Specifically in implementation, step S110 specifically includes:

[0074] Set sensors in the power supply line, power transmission and distribution line, and electrical equipment to detect electrical influence parameters in the power system, including: regional electricity consumption scale, equipment operating power, circuit switch switching frequency, and regional electricity consumption type;

[0075] Set a thermometer, clock, acoustic and optical inspection equipment in the power consumption site to collect environmental influence parameters affecting the power system load, including: temperature and humidity during equipment operation, date, noise, and light intensity.

[0076] Obtain influence parameters from each sensor at fixed intervals. After data cleaning, serialize the influence parameters in chronological order, and store the time series of each parameter in a database.

[0077] Step S120: Obtain the change function of the total circuit load over time according to the inspection frequency of the data. Decompose the change function based on the inspection frequency to obtain the component functions corresponding to the circuit load influence parameters. Calculate the integral of the component functions within the assumed period, and store the circuit load influence parameters with integrals higher than the threshold in the parameter library.

[0078] Specifically in implementation, conduct an assumed period test on the time-domain change law of the numerical values of each influence parameter to obtain the inspection frequency of the influence parameter. Decompose the load change function in the previous period according to the inspection frequency of each influence parameter, calculate the integral of the corresponding component of the influence parameter within the range of the previous period, and store the influence parameters with integral values higher than the threshold in the relevant parameter library.

[0079] In some exemplary embodiments, the conducting an assumed period test on the circuit load influence parameters to obtain the inspection frequency of the circuit load influence parameters, obtaining the change function of the total circuit load over time, decomposing the change function based on the inspection frequency to obtain the component functions corresponding to the circuit load influence parameters, calculating the integral of the component functions within the assumed period, and storing the circuit load influence parameters with integrals higher than the threshold in the parameter library includes:

[0080] Calculate the mean and variance of each element in the time series;

[0081] Perform a hypothesis period test on the time series based on the mean value and the variance to obtain the test period of the time series;

[0082] Based on the test period of the time series, obtain the test frequency of the circuit load influence parameter;

[0083] Obtain the variation function of the total circuit load over time;

[0084] Decompose the variation function based on the test frequency to obtain the component function corresponding to the circuit load influence parameter;

[0085] Calculate the integral of the component function within the assumed period, and store the circuit load influence parameters with integrals higher than the threshold in the parameter library.

[0086] In specific implementation, step S120 specifically includes:

[0087] Use the wavelet transform matrix to filter the noise in each time series, and calculate the mean value and variance of each element in the time series:

[0088]

[0089] Among them, represents the mean value of each element in the time series, S 2 represents the variance of each element in the time series, n represents the number of elements in the time series, ai represents the i-th element in the time series, and i is the sequential label;

[0090] Perform a hypothesis period test on the time series and calculate the test period of the time series:

[0091]

[0092] Among them, E represents the autocorrelation coefficient of the time series, and T is the assumed period;

[0093] Calculate the corresponding value of the assumed period T when the autocorrelation coefficient E takes the minimum value, record it as the test period of the time series, and record the reciprocal of the test period as the test frequency of the time series;

[0094] Obtain the variation function of the total circuit load over time and perform frequency domain decomposition on the function:

[0095] F[Y(t)] = Y1(f1·t) + Y2(f2·t) + … + Y m (f m ·t);

[0096] Among them, Y(t) represents the function of the total circuit load varying with time, t represents time, F[Y(t)] represents the Fourier transform of the function Y(t), m represents the total number of influencing parameters, and f1 to f m represent the test frequencies of the 1st to mth influencing parameters, and Y1 to Y m represent the component functions corresponding to the 1st to mth influencing parameters;

[0097] Determine whether the component function corresponding to the influencing parameter meets the screening condition. If it meets, store the corresponding influencing parameter as a relevant parameter in the database; otherwise, filter out the corresponding influencing parameter. The screening condition is:

[0098]

[0099] Among them, TR represents the duration of the detection period, W represents the preset screening threshold, and Y(f·t) represents the component function.

[0100] Step S130: Fit the database and the parameter library based on several methods to obtain several candidate models. According to the probability distribution of the simulation results of the candidate models, calculate the expected risk of the candidate models, and perform VC dimension analysis on the candidate models to obtain the confidence interval of the candidate models.

[0101] Specifically in implementation, different function sets are selected to fit the load data and relevant parameters to obtain candidate models. According to the probability distribution of the simulation results of the candidate models, calculate the expected risk of each candidate model, and perform VC dimension analysis on the candidate models to obtain the confidence interval of the models.

[0102] In some exemplary embodiments, the step of fitting the database and the parameter library based on several methods to obtain several candidate models, calculating the expected risk of the candidate models according to the probability distribution of the simulation results of the candidate models, and performing VC dimension analysis on the candidate models to obtain the confidence interval of the candidate models includes:

[0103] Taking the data in the database as the output quantity and the parameters in the parameter library as the input quantity, fit the database and the parameter library based on several methods to obtain several candidate models;

[0104] Calculate the simulation results of the candidate models, and calculate the expected risk of the candidate models according to the probability distribution of the simulation results of the candidate models;

[0105] Calculate the VC dimension of the candidate models to obtain the confidence interval of the candidate models.

[0106] Specifically in implementation, step S130 specifically includes:

[0107] Taking the load as the output and the relevant parameters as the inputs, different function sets are selected to fit the input-output relationship between the relevant parameters and the load. The function models are the classical fitting functions in Matlab, SVM, LSSVM or decision tree models, and the fitting result of each function model is recorded as a preselected model;

[0108] According to the prediction results of the preselected models, calculate the expected risk of the models:

[0109]

[0110] where, Remp represents the expected risk of the model, j represents the sampling serial number, T1 represents the sampling interval, L(j·T1) represents the prediction result of the preselected model in the j-th sampling, and Y(j·T1) represents the actual load obtained in the j-th sampling;

[0111] Use the relevant formulas of VC dimension theory to calculate the VC dimension of the preselected models, and obtain the confidence interval of the models as follows:

[0112]

[0113] where, h is the VC dimension of the preselected model, and α is the preset confidence level.

[0114] As an example, assume that there are 3 influencing parameters in the power system, and the sampling results within the period are [1.05, 1.2, 1.15, 0.99], [14, 20, 40, 38], and [20.1, 20.4, 21.0, 20.5] respectively, and the sampling result of the total circuit load is [200, 400, 500, 100]. Then, perform frequency decomposition on the load in the circuit for the 3 influencing factors respectively, and the closeness degrees of the load in the circuit to the influencing parameters 1, 2, and 3 are 0.18, 0.22, and 0.45 respectively. The preset screening threshold is 0.4, so the influencing parameter 3 is taken as the main parameter for model fitting.

[0115] Step S140: Screen out the preselected models whose expected risks are outside the confidence interval, perform principal component analysis on the remaining preselected models, use the function of the parameters in the parameter library corresponding to the principal components as the base kernel function of the remaining preselected models, and perform linear programming on the minimum value of the prediction error of the remaining preselected models with the value range of the base kernel function as the constraint condition.

[0116] Specifically in implementation, screen out all the preselected models whose expected risks are outside the confidence interval, perform principal component analysis on the remaining preselected models, use the function of the relevant parameters corresponding to the principal components as the base kernel function of the models, and perform linear programming on the minimum value of the model prediction error with the value range of the base kernel function as the constraint condition.

[0117] In some exemplary embodiments, for the preselected models with the expected risks outside the confidence interval, principal component analysis is performed on the remaining preselected models, and a function of the parameters of the parameter library corresponding to the principal components is used as the base kernel function of the remaining preselected models. With the value range of the base kernel function as the constraint condition, linear programming is performed on the minimum value of the prediction error of the remaining preselected models, including:

[0118] For the preselected models with the expected risks outside the confidence interval, principal component analysis is performed on the remaining preselected models, and the main function of the function set selected when fitting the remaining preselected models is used as the base kernel function of the remaining preselected models;

[0119] With the value range of the base kernel function as the constraint condition, linear programming is performed on the minimum value of the prediction error of the remaining preselected models, and the minimum value is denoted as the optimization parameter of the remaining preselected models.

[0120] In specific implementation, step S140 specifically includes:

[0121] Filter out all preselected models with expected risks not within the confidence interval, and use the main function of the function set selected when fitting the preselected models as the base kernel function of the preselected models;

[0122] With the value range of the base kernel function as the constraint condition, plan the minimum value of the prediction error of each preselected model, and the obtained minimum value is denoted as the optimization parameter of the preselected models.

[0123] Step S150: Select the remaining preselected model with the smallest prediction error as the target model, calculate the actual error volatility of the target model, and adjust the prediction result of the target model based on the actual error volatility.

[0124] In specific implementation, select the preselected model with the smallest prediction error as the target model, calculate the actual error volatility of the target model, adjust the prediction result of the target model according to the actual error volatility, and replace the prediction model when the error deviates from the preset range.

[0125] In some exemplary embodiments, the selecting the remaining preselected model with the smallest prediction error as the target model, calculating the actual error volatility of the target model, and adjusting the prediction result of the target model based on the actual error volatility includes:

[0126] Select the remaining preselected model with the smallest prediction error as the target model, and use the target model to perform load forecasting of the power system in the next cycle;

[0127] In the load forecasting process of the next cycle, detect the error between the predicted result and the measured result, generate an error sampling sequence, perform a stability analysis on the error sampling sequence, and obtain the actual error volatility of the target model;

[0128] Predict the next element value of the error sampling sequence based on the actual error volatility, and adjust the prediction result of the target model based on the next element value;

[0129] The method further includes:

[0130] In response to the actual error volatility deviating from the preset range, end the current cycle and replace the target model in the next cycle.

[0131] Specifically, step S150 specifically includes:

[0132] Select the preselected model with the smallest optimization parameter as the target model, and use the target model to perform load forecasting of the power system in the next cycle;

[0133] In the prediction process of the next cycle, detect the error between the predicted result and the measured result at fixed time intervals, generate an error sampling sequence, perform a stability analysis on the error sampling sequence, and obtain the actual error volatility of the target model;

[0134] Predict the next element value of the error sampling sequence according to the actual error volatility, adjust the prediction result of the target model according to the next element value. When the actual error volatility deviates from the preset range, end the current cycle and replace the target model in the next cycle.

[0135] As can be seen from the above, the load forecasting method provided by the embodiments of the present disclosure includes: obtaining circuit load impact parameters from a power system and the operating environment of power equipment, and storing the circuit load impact parameters in a database; performing a hypothesis period test on the circuit load impact parameters to obtain the test frequency of the circuit load impact parameters, obtaining the variation function of the total circuit load over time, decomposing the variation function based on the test frequency to obtain the component functions corresponding to the circuit load impact parameters, calculating the integral of the component functions within the hypothesis period, and storing the circuit load impact parameters with integrals higher than the threshold in a parameter library; fitting the database and the parameter library based on several methods to obtain several candidate models, calculating the expected risk of the candidate models according to the probability distribution of the simulation results of the candidate models, performing VC dimension analysis on the candidate models to obtain the confidence interval of the candidate models; screening out the candidate models with expected risks outside the confidence interval, performing principal component analysis on the remaining candidate models, using the function of the parameters in the parameter library corresponding to the principal components as the kernel function of the remaining candidate models, and performing linear programming on the minimum value of the prediction error of the remaining candidate models with the value range of the kernel function as the constraint condition; selecting the remaining candidate model with the smallest prediction error as the target model, calculating the actual error volatility of the target model, and adjusting the prediction result of the target model based on the actual error volatility.

[0136] The present disclosure improves the accuracy of load forecasting, and also improves the versatility, flexibility and timeliness. It can more accurately predict the changes in electricity consumption loads, helps to plan electricity use more effectively, avoid problems caused by the imbalance between power supply and demand, increase the accuracy of model selection, help optimize the load forecasting model, improve electricity use efficiency and the decision-making effect of the power system. Specifically:

[0137] The present disclosure can obtain multi-source parameters of electricity-consuming units for hypothesis period testing, perform frequency domain decomposition on the test period and the total load period, and use the decomposition matching results as relevant data, which can more accurately predict the changes in electricity consumption loads, helps to plan electricity use more effectively, and avoid problems caused by the imbalance between power supply and demand.

[0138] The present disclosure can perform normalization processing on relevant data one by one, select different function models to fit the load data and relevant data, calculate the expected risks of the fitting results of each model, transform the model dimensions, and discard all function models with expected risks outside the interval, thereby increasing the accuracy of model selection, better planning and allocating energy resources, and improving energy utilization efficiency.

[0139] The present disclosure uses a self-attention mechanism to calculate the main correlation dimensions in each model, takes the regression function of the main correlation function within a period as the base kernel function, constructs a linear programming function with the prediction result of the model as a constraint condition, finds the value of the optimization parameter when the optimization equation is minimized, and selects the model corresponding to the minimum optimization parameter as the target model, which can help optimize the load prediction model, thereby reasonably arranging power distribution and improving the power consumption efficiency and the decision-making effect of the power system.

[0140] It should be noted that the method of the embodiments of the present disclosure can be executed by a single device, such as a computer or a server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In this case of a distributed scenario, one of the multiple devices can only execute one or more steps of the method of the embodiments of the present disclosure, and these multiple devices will interact with each other to complete the described method.

[0141] It should be noted that some embodiments of the present disclosure have been described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the above embodiments and still achieve the desired result. Additionally, the processes depicted in the drawings do not necessarily require the specific order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0142] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present disclosure also provides a load prediction system.

[0143] Refer to Figure 2 , which is a schematic diagram of an application scenario of the load prediction system provided by an exemplary embodiment of the present disclosure.

[0144] The load prediction system includes the following modules:

[0145] A factor sensing module 210, configured to obtain circuit load impact parameters from the power system and the operating environment of power equipment, and store the circuit load impact parameters in a database;

[0146] A correlation screening module 220, configured to perform a hypothesis period test on the circuit load impact parameters to obtain the test frequency of the circuit load impact parameters, obtain the variation function of the total circuit load over time, decompose the variation function based on the test frequency to obtain the component functions corresponding to the circuit load impact parameters, calculate the integral of the component functions within the hypothesis period, and store the circuit load impact parameters with integrals higher than the threshold in a parameter library;

[0147] The function optimization module 230 is configured to fit the database and the parameter library in several ways to obtain several preliminary models, calculate the expected risk of the preliminary models according to the probability distribution of the simulation results of the preliminary models, perform VC dimension analysis on the preliminary models to obtain the confidence interval of the preliminary models;

[0148] The target regression module 240 is configured to screen out the preliminary models whose expected risk is outside the confidence interval, perform principal component analysis on the remaining preliminary models, use the function of the parameters of the parameter library corresponding to the principal components as the base kernel function of the remaining preliminary models, and perform linear programming on the minimum value of the prediction error of the remaining preliminary models with the value range of the base kernel function as the constraint condition;

[0149] The error correction module 250 is configured to select the remaining preliminary model with the smallest prediction error as the target model, calculate the actual error volatility of the target model, and adjust the prediction result of the target model based on the actual error volatility.

[0150] In some exemplary embodiments, the factor sensing module 210 is specifically configured to:

[0151] Obtain electrical influence parameters from the power system and environmental influence parameters from the operating environment of the power equipment, and the circuit load influence parameters include the electrical influence parameters and the environmental influence parameters;

[0152] Serialize the electrical influence parameters and the environmental influence parameters in chronological order to obtain the time series corresponding to the electrical influence parameters and the environmental influence parameters, and store the time series in the database.

[0153] In some exemplary embodiments, the correlation screening module 220 is specifically configured to:

[0154] Calculate the mean and variance of each element in the time series;

[0155] Perform a hypothesis periodic test on the time series based on the mean and the variance to obtain the test period of the time series;

[0156] Obtain the test frequency of the circuit load influence parameters based on the test period of the time series;

[0157] Obtain the function of the total circuit load changing with time;

[0158] Decompose the change function based on the test frequency to obtain the component function corresponding to the circuit load influence parameters;

[0159] Calculate the integral of the component function within the assumed period, and store the circuit load impact parameters with integrals higher than the threshold in the parameter library.

[0160] In some exemplary embodiments, the function optimization module 230 is specifically configured to:

[0161] Using the data in the database as the output quantity and the parameters in the parameter library as the input quantity, fit the database and the parameter library based on several methods to obtain several preliminary selection models;

[0162] Calculate the simulation results of the preliminary selection models, and calculate the expected risk of the preliminary selection models according to the probability distribution of the simulation results of the preliminary selection models;

[0163] Calculate the VC dimension of the preliminary selection models to obtain the confidence interval of the preliminary selection models.

[0164] In some exemplary embodiments, the target regression module 240 is specifically configured to:

[0165] Screen out the preliminary selection models whose expected risks are outside the confidence interval, perform principal component analysis on the remaining preliminary selection models, and use the main functions of the function set selected when fitting the remaining preliminary selection models as the base kernel functions of the remaining preliminary selection models;

[0166] Taking the value range of the base kernel function as the constraint condition, perform linear programming on the minimum value of the prediction error of the remaining preliminary selection models, and record the minimum value as the optimization parameter of the remaining preliminary selection models.

[0167] In some exemplary embodiments, the error correction module 250 is specifically configured to:

[0168] Select the remaining preliminary selection model with the smallest prediction error as the target model, and use the target model to perform load forecasting for the power system in the next period;

[0169] During the load forecasting process in the next period, detect the error between the prediction result and the measured result, generate an error sampling sequence, perform stability analysis on the error sampling sequence to obtain the actual error volatility of the target model;

[0170] Predict the next element value of the error sampling sequence based on the actual error volatility, and adjust the prediction result of the target model based on the next element value;

[0171] The method further includes:

[0172] In response to the actual error volatility deviating from the preset range, end the current period and replace the target model in the next period.

[0173] For the convenience of description, when describing the above device, various modules are described separately according to their functions. Of course, when implementing the present disclosure, the functions of each module can be implemented in one or more software and / or hardware.

[0174] The device of the above embodiment is used to implement the corresponding load prediction method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0175] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present disclosure also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the load prediction method described in any of the above embodiments.

[0176] Figure 3 FIG. shows a more specific schematic diagram of the hardware structure of the electronic device provided in this embodiment. The device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. Among them, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are communicatively connected to each other inside the device through the bus 1050.

[0177] The processor 1010 may be implemented in a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, etc., and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0178] The memory 1020 may be implemented in the form of a ROM (Read Only Memory), a RAM (Random Access Memory), a static storage device, a dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When implementing the technical solutions provided in the embodiments of this specification through software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0179] The input / output interface 1030 is used to connect to an input / output module to implement information input and output. The input / output module may be configured as a component in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Among them, the input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0180] The communication interface 1040 is used to connect to a communication module (not shown in the figure) to achieve communication and interaction between this device and other devices. The communication module can achieve communication through wired means (such as USB, network cable, etc.) or through wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0181] The bus 1050 includes a path for transmitting information between various components of the device (such as the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0182] It should be noted that although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040, and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, those skilled in the art can understand that the above device may also only include the components necessary to implement the solution of the embodiments of this specification, and does not necessarily include all the components shown in the figure.

[0183] The electronic device of the above embodiment is used to implement the corresponding load prediction method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0184] The memory 1020 stores machine-readable instructions executable by the processor 1010. When the electronic device runs, the processor 1010 communicates with the memory 1020 through the bus 1030, so that the processor 1010 executes the following instructions when running:

[0185] Obtain circuit load impact parameters from the power system and the operating environment of power equipment, and store the circuit load impact parameters in a database;

[0186] Conduct a hypothesis period test on the circuit load impact parameters to obtain the test frequency of the circuit load impact parameters, obtain the variation function of the total circuit load over time, decompose the variation function based on the test frequency to obtain the component function corresponding to the circuit load impact parameters, calculate the integral of the component function within the hypothesis period, and store the circuit load impact parameters with integrals higher than the threshold in a parameter library;

[0187] Fit the database and the parameter library based on several methods to obtain several candidate models, calculate the expected risk of the candidate models according to the probability distribution of the simulation results of the candidate models, conduct a VC dimension analysis on the candidate models to obtain the confidence interval of the candidate models;

[0188] Screen out the preselected models for which the expected risks are outside the confidence interval, perform principal component analysis on the remaining preselected models, use the function of the parameters in the parameter library corresponding to the principal components as the base kernel function of the remaining preselected models, and perform linear programming on the minimum value of the prediction error of the remaining preselected models with the value range of the base kernel function as the constraint condition;

[0189] Select the remaining preselected model with the smallest prediction error as the target model, calculate the actual error volatility of the target model, and adjust the prediction result of the target model based on the actual error volatility.

[0190] In a possible implementation manner, in the instructions executed by the processor 1010, the obtaining circuit load impact parameters from the power system and the operating environment of the power equipment and storing the circuit load impact parameters in the database includes:

[0191] Obtain electrical impact parameters from the power system and environmental impact parameters from the operating environment of the power equipment, where the circuit load impact parameters include the electrical impact parameters and the environmental impact parameters;

[0192] Serialize the electrical impact parameters and the environmental impact parameters in chronological order to obtain the time series corresponding to the electrical impact parameters and the environmental impact parameters, and store the time series in the database.

[0193] In a possible implementation manner, in the instructions executed by the processor 1010, the performing an assumed period test on the circuit load impact parameters to obtain the test frequency of the circuit load impact parameters, obtaining the variation function of the total circuit load over time, decomposing the variation function based on the test frequency to obtain the component functions corresponding to the circuit load impact parameters, and calculating the integral of the component functions within the assumed period and storing the circuit load impact parameters with the integral higher than the threshold in the parameter library includes:

[0194] Calculate the mean and variance of each element in the time series;

[0195] Perform an assumed period test on the time series based on the mean and the variance to obtain the test period of the time series;

[0196] Obtain the test frequency of the circuit load impact parameters based on the test period of the time series;

[0197] Obtain the variation function of the total circuit load over time;

[0198] Decompose the variation function based on the test frequency to obtain the component functions corresponding to the circuit load impact parameters;

[0199] Calculate the integral of the component function within the assumed period, and store the circuit load impact parameters with integrals higher than the threshold in the parameter library.

[0200] In a possible implementation, among the instructions executed by the processor 1010, the steps of fitting the database and the parameter library in several ways to obtain several candidate models, calculating the expected risk of the candidate models according to the probability distribution of the simulation results of the candidate models, and performing VC dimension analysis on the candidate models to obtain the confidence interval of the candidate models include:

[0201] Taking the data in the database as the output quantity and the parameters in the parameter library as the input quantity, fitting the database and the parameter library in several ways to obtain several candidate models;

[0202] Calculate the simulation results of the candidate models, and calculate the expected risk of the candidate models according to the probability distribution of the simulation results of the candidate models;

[0203] Calculate the VC dimension of the candidate models to obtain the confidence interval of the candidate models.

[0204] In a possible implementation, among the instructions executed by the processor 1010, the steps of screening out the candidate models with expected risks outside the confidence interval, performing principal component analysis on the remaining candidate models, using the function of the parameters in the parameter library corresponding to the principal components as the kernel function of the remaining candidate models, and performing linear programming on the minimum value of the prediction error of the remaining candidate models with the value range of the kernel function as the constraint condition include:

[0205] Screen out the candidate models with expected risks outside the confidence interval, perform principal component analysis on the remaining candidate models, and use the main function of the function set selected when fitting the remaining candidate models as the kernel function of the remaining candidate models;

[0206] With the value range of the kernel function as the constraint condition, perform linear programming on the minimum value of the prediction error of the remaining candidate models, and record the minimum value as the optimization parameter of the remaining candidate models.

[0207] In a possible implementation, among the instructions executed by the processor 1010, the steps of selecting the remaining candidate model with the smallest prediction error as the target model, calculating the actual error volatility of the target model, and adjusting the prediction result of the target model based on the actual error volatility include:

[0208] Select the remaining candidate model with the smallest prediction error as the target model, and use the target model to perform load forecasting for the power system in the next period;

[0209] In the load forecasting process of the next cycle, detect the error between the forecasting result and the measured result, generate an error sampling sequence, perform a stability analysis on the error sampling sequence, and obtain the actual error volatility of the target model;

[0210] Predict the next element value of the error sampling sequence based on the actual error volatility, and adjust the forecasting result of the target model based on the next element value;

[0211] The method further includes:

[0212] When the actual error volatility deviates from the preset range, end the current cycle and replace the target model in the next cycle.

[0213] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present disclosure also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the load forecasting method described in any one of the above embodiments.

[0214] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. The 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 technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information accessible by a computing device.

[0215] The above non-transitory computer-readable storage medium can be any available medium or data storage device accessible by a computer, including but not limited to magnetic memory (such as floppy disks, hard disks, magnetic tapes, magneto-optical discs (MO), etc.), optical memory (such as CDs, DVDs, BDs, HVDs, etc.), and semiconductor memory (such as ROM, EPROM, EEPROM, non-volatile memory (NAND FLASH), solid state drives (SSD)), etc.

[0216] The computer instructions stored in the storage medium of the above embodiment are used to cause the computer to execute the load forecasting method described in any one of the above exemplary method parts, and have the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0217] Based on the same inventive concept, corresponding to the load prediction method described in any of the above embodiments, the present disclosure also provides a computer program product, which includes computer program instructions. In some embodiments, the computer program instructions can be executed by one or more processors of a computer to cause the computer and / or the processor to execute the load prediction method. Corresponding to the execution subjects corresponding to the respective steps in the respective embodiments of the load prediction method, the processor executing the corresponding steps can belong to the corresponding execution subject.

[0218] The computer program product of the above embodiments is used to cause the computer and / or the processor to execute the load prediction method described in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be elaborated herein.

[0219] Those skilled in the art know that the embodiments of the present disclosure can be implemented as a system, a method, or a computer program product. Therefore, the present disclosure can be specifically implemented in the following forms, namely: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to as "circuit", "module" or "system" in this document. In addition, in some embodiments, the present disclosure can also be implemented in the form of a computer program product in one or more computer-readable media, which contain computer-readable program code.

[0220] Any combination of one or more computer-readable media can be adopted. The computer-readable media can be computer-readable signal media or computer-readable storage media. The computer-readable storage media can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive examples) of the computer-readable storage media can include, for example: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, the computer-readable storage media can be any tangible medium that contains or stores a program, which can be used by or in combination with an instruction execution system, apparatus, or device.

[0221] A computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, in which computer-readable program code is carried. Such a propagated data signal can take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device.

[0222] The program code contained on a computer-readable medium can be transmitted by any appropriate medium, including but not limited to wireless, wire, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0223] The computer program code for performing the operations of the present disclosure can be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0224] It should be understood that each block of the flowchart and / or block diagram, and the combinations of blocks 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, or other programmable data processing device, thereby producing a machine. These computer program instructions, when executed by a computer or other programmable data processing device, produce a device that implements the functions / operations specified in the blocks of the flowchart and / or block diagram.

[0225] These computer program instructions can also be stored in a computer-readable medium that can cause a computer or other programmable data processing device to work in a specific manner. In this way, the instructions stored in the computer-readable medium produce a product that includes an instruction device for implementing the functions / operations specified in the blocks of the flowchart and / or block diagram.

[0226] Computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other devices, so that a series of operation steps are performed on the computer, other programmable data processing apparatus, or other devices to generate a computer-implemented process, thereby enabling the instructions executed on the computer or other programmable apparatus to provide a process for implementing the functions / operations specified in the blocks of the flowchart and / or block diagram.

[0227] In addition, although the operations of the method of the present disclosure are described in a specific order in the drawings, this does not require or imply that the operations must be performed in that specific order, or that all of the illustrated operations must be performed to achieve the desired result. On the contrary, the steps depicted in the flowchart can be changed in the order of execution. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step for execution, and / or one step may be decomposed into multiple steps for execution.

[0228] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. Among them, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the above-mentioned module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutively represented blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, as well as the combination of blocks in the block diagram or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0229] It should be noted that although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of the two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0230] Those of ordinary skill in the art should understand that any discussion of the above embodiments is merely exemplary and is not intended to imply that the scope of the present application (including the claims) is limited to these examples; within the concept of the present application, the technical features in the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations in different aspects of the embodiments of the present application as described above, and for the sake of brevity, they are not provided in detail.

[0231] In addition, for simplicity of explanation and discussion, and so as not to make the embodiments of the present application difficult to understand, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Further, the devices may be shown in block diagram form so as not to make the embodiments of the present application difficult to understand, and this also takes into account the fact that the details of the implementation of such block diagram devices are highly dependent on the platform on which the embodiments of the present application are to be implemented (i.e., these details should be entirely within the understanding of those skilled in the art). In cases where specific details (such as circuits) are set forth to describe exemplary embodiments of the present application, it will be apparent to those skilled in the art that the embodiments of the present application can be implemented without these specific details or with variations of these specific details. Accordingly, these descriptions should be considered illustrative rather than restrictive.

[0232] Although the present application has been described in connection with specific embodiments of the present application, many alternatives, modifications, and variations of these embodiments will be apparent to those of ordinary skill in the art based on the foregoing description. For example, other memory architectures (such as dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0233] The embodiments of the present application are intended to cover all such alternatives, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of the present application shall be included within the protection scope of the present application.

[0234] Although the spirit and principles of the present disclosure have been described with reference to several specific embodiments, it should be understood that the present disclosure is not limited to the specific embodiments disclosed, and the division of each aspect does not mean that the features in these aspects cannot be combined for benefit, and this division is only for the convenience of expression. The present disclosure is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. The scope of the appended claims is to be accorded the broadest interpretation so as to encompass all such modifications and equivalent structures and functions.

Claims

1. A load forecasting method, characterized in that: include: Acquire circuit load impact parameters from the power system and power equipment operating environment, and store the circuit load impact parameters in a database; Performing a hypothesis period test on the circuit load influencing parameter to obtain a test frequency of the circuit load influencing parameter, obtaining a function of the total circuit load changing over time, decomposing the change function based on the test frequency to obtain a component function corresponding to the circuit load influencing parameter, calculating the integral of the component function within the hypothesis period, and storing the circuit load influencing parameter having an integral higher than a threshold in a parameter library; Fitting the database and the parameter library based on several methods to obtain several pre-selected models, calculating the expected risk of the pre-selected models according to the probability distribution of the simulation results of the pre-selected models, performing VC dimension analysis on the pre-selected models, and obtaining the confidence interval of the pre-selected models; Screening out the pre-selected models whose expected risks are outside the confidence interval, performing principal component analysis on the remaining pre-selected models, using the function of the parameters of the parameter library corresponding to the principal components as the base kernel function of the remaining pre-selected models, and using the value range of the base kernel function as a restriction condition to perform linear programming on the minimum value of the prediction error of the remaining pre-selected models; The remaining pre-selected model with the smallest prediction error is selected as the target model, the actual error volatility of the target model is calculated, and the prediction result of the target model is adjusted based on the actual error volatility.

2. The method according to claim 1, characterized in that The obtaining of circuit load influence parameters from the power system and the power equipment operating environment and storing the circuit load influence parameters in a database includes: Acquiring electrical influencing parameters from the power system and acquiring environmental influencing parameters from the power equipment operating environment, wherein the circuit load influencing parameters include the electrical influencing parameters and the environmental influencing parameters; The electrical influencing parameters and the environmental influencing parameters are serialized in time order to obtain time series corresponding to the electrical influencing parameters and the environmental influencing parameters, and the time series are stored in a database.

3. The method according to claim 2, characterized in that The circuit load influencing parameter is subjected to a hypothesis period test to obtain a test frequency of the circuit load influencing parameter, obtain a function of a change of the total circuit load over time, decompose the change function based on the test frequency to obtain a component function corresponding to the circuit load influencing parameter, calculate an integral of the component function within the hypothesis period, and store the circuit load influencing parameter having an integral higher than a threshold into a parameter library, including: Calculate the mean and variance of each element in the time series; Performing a hypothesis period test on the time series based on the mean and the variance to obtain a test period of the time series; Based on the inspection period of the time series, obtaining the inspection frequency of the circuit load influencing parameter; Obtain the function of total circuit load changing with time; Decomposing the variation function based on the inspection frequency to obtain a component function corresponding to the circuit load influencing parameter; The integral of the component function within the assumed period is calculated, and the circuit load influencing parameters whose integral is higher than a threshold are stored in a parameter library.

4. The method according to claim 1, characterized in that: The method of fitting the database and the parameter library based on several methods to obtain several pre-selected models, calculating the expected risk of the pre-selected models according to the probability distribution of the simulation results of the pre-selected models, and performing VC dimension analysis on the pre-selected models to obtain the confidence interval of the pre-selected models includes: Using the data in the database as output and the parameters in the parameter library as input, fitting the database and the parameter library based on several methods to obtain several pre-selected models; Calculating the simulation results of the pre-selected model, and calculating the expected risk of the pre-selected model according to the probability distribution of the simulation results of the pre-selected model; The VC dimension of the pre-selected model is calculated to obtain the confidence interval of the pre-selected model.

5. The method according to claim 1, characterized in that The method of screening out the pre-selected models whose expected risk is outside the confidence interval, performing principal component analysis on the remaining pre-selected models, using the function of the parameters of the parameter library corresponding to the principal component as the base kernel function of the remaining pre-selected models, and using the value range of the base kernel function as a restriction condition to perform linear programming on the minimum value of the prediction error of the remaining pre-selected models, includes: Screening out the pre-selected models whose expected risks are outside the confidence interval, performing principal component analysis on the remaining pre-selected models, and using the principal function of the function set selected when fitting the remaining pre-selected models as the base kernel function of the remaining pre-selected models; Taking the value range of the base kernel function as a constraint condition, linear programming is performed on the minimum value of the prediction error of the remaining pre-selected model, and the minimum value is recorded as the optimization parameter of the remaining pre-selected model.

6. The method according to claim 1, characterized in that The selecting the remaining pre-selected model with the smallest prediction error as the target model, calculating the actual error volatility of the target model, and adjusting the prediction result of the target model based on the actual error volatility includes: Selecting the remaining pre-selected model with the smallest prediction error as the target model, and using the target model to perform load forecasting for the power system in the next cycle; In the load forecasting process of the next cycle, the error between the forecast result and the measured result is detected, an error sampling sequence is generated, and a stability analysis is performed on the error sampling sequence to obtain the actual error volatility of the target model; Predicting a next element value of the error sampling sequence based on the actual error volatility, and adjusting a prediction result of the target model based on the next element value; The method further comprises: In response to the actual error fluctuation rate deviating from a preset range, the current cycle is ended, and the target model is replaced in the next cycle.

7. A load forecasting system, characterized in that: include: A factor sensing module is configured to obtain circuit load influence parameters from the power system and the power equipment operating environment, and store the circuit load influence parameters in a database; A related screening module is configured to perform a hypothesis period test on the circuit load influencing parameter, obtain a test frequency of the circuit load influencing parameter, obtain a function of the change of the total circuit load over time, decompose the change function based on the test frequency, obtain a component function corresponding to the circuit load influencing parameter, calculate an integral of the component function within the hypothesis period, and store the circuit load influencing parameter with an integral higher than a threshold in a parameter library; A function optimization module is configured to fit the database and the parameter library based on several methods to obtain several pre-selected models, calculate the expected risk of the pre-selected models according to the probability distribution of the simulation results of the pre-selected models, perform VC dimension analysis on the pre-selected models, and obtain the confidence interval of the pre-selected models; a target regression module, configured to screen out the pre-selected models whose expected risk is outside the confidence interval, perform principal component analysis on the remaining pre-selected models, use the function of the parameters of the parameter library corresponding to the principal component as the base kernel function of the remaining pre-selected models, use the value range of the base kernel function as a restriction condition, and perform linear programming on the minimum value of the prediction error of the remaining pre-selected models; The error correction module is configured to select the remaining pre-selected model with the smallest prediction error as the target model, calculate the actual error volatility of the target model, and adjust the prediction result of the target model based on the actual error volatility.

8. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method according to any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to cause a computer to execute the method according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises computer program instructions, which, when executed on a computer, cause the computer to execute the method according to any one of claims 1 to 6.

Images