CN112734141A - Diversified load interval prediction method and device - Google Patents
Diversified load interval prediction method and device Download PDFInfo
- Publication number
- CN112734141A CN112734141A CN202110211746.7A CN202110211746A CN112734141A CN 112734141 A CN112734141 A CN 112734141A CN 202110211746 A CN202110211746 A CN 202110211746A CN 112734141 A CN112734141 A CN 112734141A
- Authority
- CN
- China
- Prior art keywords
- prediction
- load
- interval
- error
- diversified
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 50
- 238000005457 optimization Methods 0.000 claims abstract description 19
- 238000005315 distribution function Methods 0.000 claims description 31
- 238000011156 evaluation Methods 0.000 claims description 29
- 230000006870 function Effects 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 9
- 230000035945 sensitivity Effects 0.000 claims description 9
- 239000002131 composite material Substances 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 4
- 238000010586 diagram Methods 0.000 description 8
- 238000000354 decomposition reaction Methods 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 2
- 238000000513 principal component analysis Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Water Supply & Treatment (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The application discloses a diversified load interval prediction method and device. Then dividing the load predicted value into different prediction sections, fitting the error distribution of the load predicted value in each prediction section by normal distribution, and establishing a diversified load condition probability section prediction model. And then, carrying out iterative optimization on the diversified load conditional probability interval prediction model, and finally determining and evaluating the optimal prediction interval of the diversified load. The method and the device consider that only point prediction is carried out on the diversified loads, inevitable prediction errors exist, and therefore interval prediction is carried out on the diversified loads, so that the prediction result is more accurate, and practical and reliable information is provided for operation regulation and control of the power system.
Description
Technical Field
The application relates to the technical field of load prediction and planning, in particular to a diversified load interval prediction method and device.
Background
The load prediction is to determine load data of a certain future moment according to various factors such as the operating characteristics, capacity increase decision, natural conditions and social influence of a system under the condition of meeting a certain precision requirement, wherein the load refers to the power demand (power) or the power consumption. Load prediction is an important content in economic dispatch of a power system and is an important module of an Energy Management System (EMS).
In the prior art, the future power demand of a power load is presumed through the past power demand and the present power demand of the power load, specifically, point prediction is carried out on related data information of the load to obtain a predicted value of the load in a future period of time, and then starting and stopping of a generator set in a power grid are economically and reasonably arranged through the predicted value, so that the safety and stability of power grid operation are kept, a unit maintenance plan is reasonably arranged, the power generation cost is effectively reduced, and the economic benefit is improved.
However, with the rapid development of the modern society, the types and forms of loads are being developed in various directions, including specifically, power loads, thermal loads, and the like. Compared with the traditional power load, the novel diversified load is more complex and diversified, the requirements of different types of loads on the reliability and the power supply quality of a power system are different, only the point prediction is carried out on the diversified load, and the accuracy is very low.
Disclosure of Invention
The application provides a diversified load interval prediction method and device, which aim to solve the technical problem that in the prior art, only point prediction is carried out on diversified loads, and the accuracy is low.
A diversified load interval prediction method, comprising:
acquiring diversified load historical data information and multiple groups of load real values, performing point prediction on the diversified load historical data information, and determining multiple groups of load predicted values corresponding to the multiple groups of load real values one by one;
determining the error of each group of load predicted values according to the plurality of groups of load predicted values and the plurality of groups of load real values which correspond one to one;
acquiring a plurality of groups of prediction error normal distribution functions, wherein each group of prediction error normal distribution functions corresponds to the errors of each group of load predicted values one by one;
according to the multiple groups of load predicted values and preset predicted interval boundary parameters, carrying out interval division on the multiple groups of load predicted values, and determining multiple predicted intervals;
generating a diversified load conditional probability interval prediction model according to the multiple groups of prediction error normal distribution functions and the multiple prediction intervals;
performing iterative optimization on the diversified load condition probability interval prediction model, updating the boundary parameter of the prediction interval in the optimization process, and determining the optimal boundary parameter value of the prediction interval; the optimal boundary parameter value is a prediction interval boundary parameter value which enables the range of each prediction interval to be minimum on the premise of ensuring that the real value of each group of loads falls in the corresponding prediction interval;
and determining the optimal prediction interval of the diversified loads according to the optimal prediction interval boundary parameter.
Optionally, the obtaining multiple groups of prediction error normal distribution functions includes:
determining an error mean value and an error standard deviation corresponding to the error of each group of load predicted values according to the error of each group of load predicted values;
and determining a plurality of groups of prediction error normal distribution functions according to the error of each group of load predicted values, the error mean value and the error standard deviation.
Optionally, the performing interval division on the multiple groups of load predicted values according to the multiple groups of load predicted values and preset prediction interval boundary parameters, and determining multiple prediction intervals includes:
determining a normal distribution quantile corresponding to the boundary parameter of the prediction interval according to the preset boundary parameter of the prediction interval;
determining an error mean value and an error standard deviation corresponding to the error of each group of load predicted values according to the error of each group of load predicted values;
for any load predicted value, determining the upper limit value and the lower limit value of a group of predicted sections according to the normal distribution quantiles, the error mean value and the error standard deviation;
determining a boundary value of a group of prediction intervals according to the boundary parameter of the prediction intervals and the upper and lower limit values of the group of prediction intervals;
determining a predicted interval according to the upper and lower limit values of the group of predicted intervals and the boundary value of the group of predicted intervals;
a plurality of predicted block segments are obtained.
Optionally, after determining the optimal prediction interval of the diversified loads, the method further includes:
performing prediction evaluation on the optimal prediction interval; the prediction evaluation is used for verifying the validity and correctness of the optimal prediction interval.
Optionally, the prediction interval evaluation includes: average coverage error assessment, sensitivity assessment, and composite score value assessment.
A second aspect of the present application discloses a diversified load section prediction apparatus, which is applied to the diversified load section prediction method according to the first aspect, and the apparatus includes:
the point prediction module is used for acquiring diversified load historical data information and multiple groups of load real values, performing point prediction on the diversified load historical data information and determining multiple groups of load predicted values which are in one-to-one correspondence with the multiple groups of load real values;
the error acquisition module is used for determining the error of each group of load predicted values according to the one-to-one corresponding groups of load predicted values and the groups of load real values;
the function acquisition module is used for acquiring a plurality of groups of prediction error normal distribution functions, and each group of prediction error normal distribution functions corresponds to the error of each group of load predicted values one by one;
the interval division module is used for carrying out interval division on the multiple groups of load predicted values according to the multiple groups of load predicted values and preset predicted interval boundary parameters and determining multiple predicted intervals;
the interval prediction model establishing module is used for generating a diversified load condition probability interval prediction model according to the multiple groups of prediction error normal distribution functions and the multiple prediction intervals;
the optimization processing module is used for performing iterative optimization on the diversified load condition probability interval prediction model, updating the boundary parameter of the prediction interval in the optimization process and determining the optimal boundary parameter value of the prediction interval; the optimal boundary parameter value is a prediction interval boundary parameter value which enables the range of each prediction interval to be minimum on the premise of ensuring that the real value of each group of loads falls in the corresponding prediction interval;
and the optimal prediction interval acquisition module is used for determining the optimal prediction interval of the diversified loads according to the boundary parameter of the optimal prediction interval.
Optionally, the function obtaining module includes:
the mean standard deviation obtaining unit is used for determining an error mean value and an error standard deviation corresponding to the error of each group of load predicted values according to the error of each group of load predicted values;
and the function acquisition unit is used for determining a plurality of groups of prediction error normal distribution functions according to the error of each group of load prediction values, the error mean value and the error standard deviation.
Optionally, the interval dividing module includes:
a normal distribution quantile obtaining unit, configured to determine, according to the preset prediction interval boundary parameter, a normal distribution quantile corresponding to the prediction interval boundary parameter;
the mean standard deviation obtaining unit is used for determining an error mean value and an error standard deviation corresponding to the error of each group of load predicted values according to the error of each group of load predicted values;
the interval upper and lower limit value acquisition unit is used for determining the upper and lower limit values of a group of prediction interval values according to the normal distribution quantile points, the error mean values and the error standard deviations aiming at any load prediction value;
a boundary value obtaining unit, configured to determine a boundary value of a group of predicted sections according to the predicted section boundary parameter and upper and lower limit values of the group of predicted sections;
a predicted block section obtaining unit configured to determine a predicted block section based on upper and lower limit values of the group of predicted block sections and a boundary value of the group of predicted block sections; a plurality of predicted block segments are obtained.
Optionally, the diversified load interval prediction apparatus further includes a prediction evaluation module, where the prediction evaluation module is configured to perform prediction evaluation on the optimal prediction interval after determining the optimal prediction interval of the diversified load; the prediction evaluation is used for verifying the validity and correctness of the optimal prediction interval.
Optionally, the prediction interval evaluation includes: average coverage error assessment, sensitivity assessment, and composite score value assessment.
The application discloses a diversified load interval prediction method and device. Then dividing the load predicted value into different prediction sections, fitting the error distribution of the load predicted value in each prediction section by normal distribution, and establishing a diversified load condition probability section prediction model. And then, carrying out iterative optimization on the diversified load conditional probability interval prediction model, and finally determining the optimal prediction interval of the diversified load. The effectiveness and the correctness of the optimal prediction interval are verified by performing prediction evaluation on the optimal prediction interval.
In the present application, it is considered that only point prediction is performed on diversified loads, and there is an inevitable prediction error, and therefore it is important to obtain the upper and lower limit values of prediction. Therefore, interval prediction is carried out on diversified loads, and on the premise that the true value of each group of loads is ensured to fall into the correspondingly obtained prediction interval, the smaller the range of each prediction interval is, the better the prediction result is, so that the prediction result is more accurate, and practical and reliable information is provided for operation regulation and control of the power system.
Drawings
In order to more clearly explain the technical solution of the present application, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious to those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flowchart of a diversified load interval prediction method disclosed in an embodiment of the present application;
fig. 2 is a schematic diagram illustrating a prediction interval division in a diversified load interval prediction method disclosed in an embodiment of the present application;
fig. 3 is an error distribution histogram of a load predicted value in a diversified load interval prediction method disclosed in the embodiment of the present application;
fig. 4 is a diagram of a prediction result of an optimal prediction interval when a boundary parameter of the prediction interval is 0.95 in the diversified load interval prediction method disclosed in the embodiment of the present application;
fig. 5 is a schematic structural diagram of a diversified load interval prediction apparatus disclosed in an embodiment of the present application.
Detailed Description
In order to solve the technical problem that in the prior art, the accuracy is low only by performing point prediction on diversified loads, the application discloses a diversified load interval prediction method and device through the following two embodiments.
A first embodiment of the present application discloses a diversified load interval prediction method, specifically referring to a flow diagram shown in fig. 1, the method includes:
step S101, obtaining diversified load historical data information and multiple groups of load real values, performing point prediction on the diversified load historical data information, and determining multiple groups of load predicted values corresponding to the multiple groups of load real values one by one.
Specifically, there are many point prediction methods, including a method for establishing load prediction through machine learning, and a prediction method for mining intrinsic characteristics of load data through deep learning, including a time series method, a persistence method, a gray prediction method, a support vector machine, a neural network algorithm, and the like. The point prediction method has inevitable prediction errors, so that the obtaining of the upper limit value and the lower limit value of the prediction becomes more important, the error of the prediction result is caused by a plurality of reasons, and the error distribution functions obeyed by the prediction errors have differences under different conditions.
In some embodiments of the present application, point prediction of diversified loads is performed by taking wavelet decomposition and LSTM network in machine learning as an example, and a point prediction error condition is obtained. Specifically, in the actual operation process, 10 months of diversified load historical data information is collected, the Pearson correlation coefficient value between each feature and diversified loads at the future four-hour moment is calculated, and strong correlation features are screened to serve as a data set. And performing wavelet decomposition on the diversified load historical data to obtain a plurality of wavelet components. And establishing an input and output model of each wavelet decomposition by taking a plurality of power characteristics before the current moment as input variables and taking the load predicted value as an output variable. And (3) carrying out 0-mean normalization processing on the whole sample set, carrying out PCA (principal component analysis) dimensionality reduction on the input variable set, dividing 3000 samples as a training set of the model, and dividing 6000 samples as a test set. And carrying out LSTM network training on all wavelet component sample training sets, substituting input variables of respective wavelet sample testing sets into the trained network model to obtain predicted values and real values of all wavelet samples, and overlapping the predicted values and the real values of all wavelet components to finish the diversified load point prediction based on wavelet decomposition and the LSTM network.
And S102, determining the error of each group of load predicted values according to the multiple groups of load predicted values and the multiple groups of load real values which correspond one to one.
And step S103, acquiring multiple groups of prediction error normal distribution functions, wherein each group of prediction error normal distribution functions corresponds to the error of each group of load predicted values one by one.
Further, the obtaining of the multiple groups of prediction error normal distribution functions includes:
and determining an error mean value and an error standard deviation corresponding to the errors of each group of load predicted values according to the errors of each group of load predicted values.
And determining a plurality of groups of prediction error normal distribution functions according to the error of each group of load predicted values, the error mean value and the error standard deviation.
Specifically, the error between the predicted value of the load and the true value of the load is used as a condition, and parameterized distribution functions of prediction errors in different predicted value intervals are constructed so as to be used for describing the prediction errors of diversified loads. In some embodiments of the present application, a large amount of collected diversified load historical data information and a load true value are analyzed to obtain an error of a load predicted value, and then, an analysis modeling is performed, and a random variable of the error of the load predicted value is denoted as X, so that a probability density function of X is expressed in the following form:
in the formula: mu and sigma are respectively the mean value and the standard deviation of the error X of the load predicted value.
The distribution function corresponding to the error of the load predicted value is expressed as:
and step S104, according to the multiple groups of load predicted values and preset predicted interval boundary parameters, carrying out interval division on the multiple groups of load predicted values, and determining multiple predicted intervals.
Further, the performing interval division on the multiple groups of load predicted values according to the multiple groups of load predicted values and preset prediction interval boundary parameters, and determining multiple prediction intervals includes:
and determining the normal distribution quantile corresponding to the boundary parameter of the prediction interval according to the preset boundary parameter of the prediction interval.
And determining an error mean value and an error standard deviation corresponding to the errors of each group of load predicted values according to the errors of each group of load predicted values.
And determining the upper limit value and the lower limit value of a group of prediction sections according to the normal distribution quantiles, the error mean value and the error standard deviation aiming at any load prediction value.
And determining the boundary value of a group of prediction sections according to the boundary parameter of the prediction sections and the upper and lower limit values of the group of prediction sections.
And determining a predicted interval according to the upper and lower limit values of the group of predicted intervals and the boundary value of the group of predicted intervals.
Thereby obtaining a plurality of predicted sections.
Referring to fig. 2, a schematic diagram of prediction section division in the diversified load section prediction method disclosed in the embodiment of the present application is shown. Specifically, an initial value [ α ] of the prediction section boundary parameter is set1，α2，,α3，α4，...，αn]If the boundary parameter α of the prediction interval is used to divide the prediction interval, the error of the predicted load value can be calculated to be [ min, t [ ]1]，[t1，t2]，[t2，t3]，[t3，t4]，[t4，t5]，…，[tn，max]And predicting a normal distribution model in the interval.
According to the theory of probability statistics, under the condition of normal distribution and the condition of a prediction interval boundary parameter alpha for evaluating confidence coefficient, the confidence interval of probability prediction of diversified loads is as follows:
max＝y+z*σ+u
min＝y-z*σ+u
in the formula: max is the upper limit value of the prediction interval, min is the lower limit value of the prediction interval, y is the prediction value, z is the normal distribution quantile point corresponding to the boundary parameter alpha of the prediction interval, and sigma and u are the standard deviation and the mean value of the error of the load prediction value.
Boundary value t of predicted value intervaliAnd a prediction interval boundary parameter alphaiThe relationship between them is as follows:
ti＝min+αi*(max-min)
in the formula: alpha is alphaiIs a constant between 0 and 1, which determines the quality of the final divided segment.
Specifically, the conditional probability interval prediction is a more objective and reliable probability prediction method established by dividing prediction errors into different modules according to a certain characteristic, so that the diversified load prediction values are divided into different prediction blocks, and the distribution function of the diversified load prediction errors in each block is fitted into normal distribution, so that the error distribution description of the load prediction values can be more exquisite, and the final probability prediction result is more reliable.
And step S105, generating a diversified load condition probability interval prediction model according to the multiple groups of prediction error normal distribution functions and the multiple prediction intervals.
And S106, performing iterative optimization on the diversified load condition probability interval prediction model, and updating the boundary parameter of the prediction interval in the optimization process to determine the optimal boundary parameter value of the prediction interval. The optimal boundary parameter value is a prediction interval boundary parameter value which enables the range of each prediction interval to be minimum on the premise of ensuring that the real value of each group of loads falls in the corresponding prediction interval.
In some embodiments of the application, a multivariate load condition probability interval prediction model is optimized based on a PSO optimization method, and an optimal prediction interval boundary parameter value is searched, so that the comprehensive quality of a prediction interval is optimal.
And S107, determining the optimal prediction interval of the diversified loads according to the optimal prediction interval boundary parameter.
Further, after the determining the optimal prediction interval of the diversified loads, the method further includes:
and performing prediction evaluation on the optimal prediction intervals of a plurality of samples. The prediction evaluation is used for verifying the validity and correctness of the optimal prediction interval. The prediction interval evaluation comprises: average coverage error assessment, sensitivity assessment, and composite score value assessment.
Specifically, the average coverage error evaluation is used for evaluating the reliability of the optimal prediction interval, and the smaller the absolute value of the average coverage error evaluation is, the more reliable the prediction interval is, specifically, the following formula is used for performing:
R＝[ζ-(1-α)]×100％
wherein, R represents the average coverage rate error of the optimal prediction interval when the boundary parameter of the prediction interval is 1-alpha, and ζ represents the actual coverage rate of the optimal prediction interval when the boundary parameter of the prediction interval is 1-alpha.
And sensitivity evaluation is used for describing the change situation of the average width of the optimal prediction interval, and the lower the sensitivity value is, the narrower the interval width is, and the sensitivity evaluation is specifically carried out by the following formula:
in the formula, n represents the number of test samples, upiAnd lowiRespectively represents the upper limit value and the lower limit value of the optimal prediction interval of the ith sample, and R represents the maximum capacity value of the load.
And evaluating the comprehensive score value, which is used for determining the comprehensive quality of the optimal prediction interval, wherein the smaller the absolute value of the comprehensive score value is, the higher the quality of the optimal prediction interval is, and otherwise, the worse the quality is, and the method is specifically carried out by the following formula:
in the formula (I), the compound is shown in the specification,
the optimal prediction interval width of the ith sample is shown when the boundary parameter of the prediction interval is 1-alpha, up (i) and low (i) respectively show the optimal prediction interval upper limit value and lower limit value of the ith sample, yt (i) shows the real value of the ith sample, n shows the number of test samples, sc (i) shows the score of the ith sample, and S shows the comprehensive score value of the optimal prediction interval.
In the actual operation process, specifically referring to fig. 3, it is a distribution diagram of errors of the predicted values of the load in the diversified load interval prediction method disclosed in the embodiment of the present application, wherein the graphs (a), (b), (c) and (d) in fig. 3 respectively show prediction error histograms of the predicted values of the load with errors of [0, 3.701], [3.701, 22.952], [22.952, 32.984] and [32.984, 148.5 ]. Specifically, referring to fig. 4, it is a prediction result diagram of an optimal prediction interval when a prediction interval boundary parameter is 0.95 in the diversified load interval prediction method disclosed in the embodiment of the present application. According to the graph, the conditional normal distribution model is introduced into the diversified load interval prediction, so that the reliability of the probability prediction result can be improved, the quality of the prediction interval can be improved, and reliable information and basis can be provided for the operation and regulation of the system.
According to the technical scheme, the method and the device for predicting the diversified load interval are disclosed. Then dividing the load predicted value into different prediction sections, fitting the error distribution of the load predicted value in each prediction section by normal distribution, and establishing a diversified load condition probability section prediction model. And then, carrying out iterative optimization on the diversified load conditional probability interval prediction model, and finally determining the optimal prediction interval of the diversified load. The effectiveness and the correctness of the optimal prediction interval are verified by performing prediction evaluation on the optimal prediction interval.
In the actual operation process, the point prediction is only carried out on diversified loads, inevitable prediction errors exist, and therefore the upper limit value and the lower limit value of the prediction are important to obtain. Therefore, interval prediction is carried out on diversified loads, and on the premise that the true value of each group of loads is ensured to fall into the correspondingly obtained prediction interval, the smaller the range of each prediction interval is, the better the prediction result is, so that the prediction result is more accurate, and practical and reliable information is provided for operation regulation and control of the power system.
The following are embodiments of the apparatus disclosed herein for performing the above-described method embodiments. For details not disclosed in the device embodiments, refer to the method embodiments.
A second embodiment of the present application discloses a diversified load interval prediction apparatus, which is applied to the diversified load interval prediction method according to the first aspect, and with reference to a schematic structural diagram shown in fig. 5, the apparatus includes:
and the point prediction module 10 is configured to obtain diversified load historical data information and multiple groups of load real values, perform point prediction on the diversified load historical data information, and determine multiple groups of load predicted values corresponding to the multiple groups of load real values one to one.
And the error obtaining module 20 is configured to determine an error of each group of load predicted values according to the one-to-one corresponding multiple groups of load predicted values and multiple groups of load true values.
And the function obtaining module 30 is configured to obtain multiple groups of prediction error normal distribution functions, where each group of prediction error normal distribution functions corresponds to errors of each group of load predicted values one to one.
Further, the function obtaining module 30 includes:
and the mean standard deviation obtaining unit is used for determining an error mean value and an error standard deviation corresponding to the error of each group of load predicted values according to the error of each group of load predicted values.
And the function acquisition unit is used for determining a plurality of groups of prediction error normal distribution functions according to the error of each group of load prediction values, the error mean value and the error standard deviation.
And the interval division module 40 is configured to perform interval division on the multiple groups of load predicted values according to the multiple groups of load predicted values and preset predicted interval boundary parameters, and determine multiple predicted intervals.
Further, the interval dividing module 40 includes:
and the normal distribution quantile acquisition unit is used for determining the normal distribution quantile corresponding to the boundary parameter of the prediction interval according to the preset boundary parameter of the prediction interval.
And the mean standard deviation obtaining unit is used for determining an error mean value and an error standard deviation corresponding to the error of each group of load predicted values according to the error of each group of load predicted values.
And the interval upper and lower limit value acquisition unit is used for determining the upper and lower limit values of a group of prediction interval values according to the normal distribution quantile points, the error mean values and the error standard deviations aiming at any load prediction value.
And the boundary value acquisition unit is used for determining the boundary value of a group of predicted intervals according to the predicted interval boundary parameter and the upper and lower limit values of the group of predicted intervals.
And the predicted block section acquisition unit is used for determining a predicted block section according to the upper limit value and the lower limit value of the group of predicted block sections and the boundary value of the group of predicted block sections.
A plurality of predicted block segments are obtained.
And the interval prediction model establishing module 50 is configured to generate a diversified load condition probability interval prediction model according to the plurality of groups of prediction error normal distribution functions and the plurality of prediction intervals.
And the optimization processing module 60 is configured to perform iterative optimization on the diversified load condition probability interval prediction model, and update the prediction interval boundary parameter in the optimization process to determine an optimal prediction interval boundary parameter value. The optimal boundary parameter value is a prediction interval boundary parameter value which enables the range of each prediction interval to be minimum on the premise of ensuring that the real value of each group of loads falls in the corresponding prediction interval.
And an optimal prediction interval obtaining module 70, configured to determine an optimal prediction interval of the diversified loads according to the optimal prediction interval boundary parameter.
Further, the diversified load interval prediction device further comprises a prediction evaluation module, and the prediction evaluation module is used for performing prediction evaluation on the optimal prediction interval after the optimal prediction interval of the diversified load is determined. The prediction evaluation is used for verifying the validity and correctness of the optimal prediction interval.
Further, the prediction interval evaluation comprises: average coverage error assessment, sensitivity assessment, and composite score value assessment.
The present application has been described in detail with reference to specific embodiments and illustrative examples, but the description is not intended to limit the application. Those skilled in the art will appreciate that various equivalent substitutions, modifications or improvements may be made to the presently disclosed embodiments and implementations thereof without departing from the spirit and scope of the present disclosure, and these fall within the scope of the present disclosure. The protection scope of this application is subject to the appended claims.
Claims (10)
1. A diversified load interval prediction method, comprising:
acquiring diversified load historical data information and multiple groups of load real values, performing point prediction on the diversified load historical data information, and determining multiple groups of load predicted values corresponding to the multiple groups of load real values one by one;
determining the error of each group of load predicted values according to the plurality of groups of load predicted values and the plurality of groups of load real values which correspond one to one;
acquiring a plurality of groups of prediction error normal distribution functions, wherein each group of prediction error normal distribution functions corresponds to the errors of each group of load predicted values one by one;
according to the multiple groups of load predicted values and preset predicted interval boundary parameters, carrying out interval division on the multiple groups of load predicted values, and determining multiple predicted intervals;
generating a diversified load conditional probability interval prediction model according to the multiple groups of prediction error normal distribution functions and the multiple prediction intervals;
performing iterative optimization on the diversified load condition probability interval prediction model, updating the boundary parameter of the prediction interval in the optimization process, and determining the optimal boundary parameter value of the prediction interval; the optimal boundary parameter value is a prediction interval boundary parameter value which enables the range of each prediction interval to be minimum on the premise of ensuring that the real value of each group of loads falls in the corresponding prediction interval;
and determining the optimal prediction interval of the diversified loads according to the optimal prediction interval boundary parameter.
2. The diversified load interval prediction method according to claim 1, wherein said obtaining a plurality of groups of prediction error normal distribution functions comprises:
determining an error mean value and an error standard deviation corresponding to the error of each group of load predicted values according to the error of each group of load predicted values;
and determining a plurality of groups of prediction error normal distribution functions according to the error of each group of load predicted values, the error mean value and the error standard deviation.
3. The diversified load interval prediction method according to claim 1, wherein the interval-dividing the multiple groups of load prediction values according to the multiple groups of load prediction values and a preset prediction interval boundary parameter and determining multiple prediction intervals comprises:
determining a normal distribution quantile corresponding to the boundary parameter of the prediction interval according to the preset boundary parameter of the prediction interval;
determining an error mean value and an error standard deviation corresponding to the error of each group of load predicted values according to the error of each group of load predicted values;
for any load predicted value, determining the upper limit value and the lower limit value of a group of predicted sections according to the normal distribution quantiles, the error mean value and the error standard deviation;
determining a boundary value of a group of prediction intervals according to the boundary parameter of the prediction intervals and the upper and lower limit values of the group of prediction intervals;
determining a predicted interval according to the upper and lower limit values of the group of predicted intervals and the boundary value of the group of predicted intervals;
a plurality of predicted block segments are obtained.
4. The diversified load interval prediction method according to claim 1, further comprising, after said determining an optimal prediction interval of the diversified load:
performing prediction evaluation on the optimal prediction interval; the prediction evaluation is used for verifying the validity and correctness of the optimal prediction interval.
5. The diversified load interval prediction method according to claim 4, wherein said prediction interval evaluation comprises: average coverage error assessment, sensitivity assessment, and composite score value assessment.
6. A diversified load section prediction apparatus applied to the diversified load section prediction method according to any one of claims 1 to 5, the apparatus comprising:
the point prediction module is used for acquiring diversified load historical data information and multiple groups of load real values, performing point prediction on the diversified load historical data information and determining multiple groups of load predicted values which are in one-to-one correspondence with the multiple groups of load real values;
the error acquisition module is used for determining the error of each group of load predicted values according to the one-to-one corresponding groups of load predicted values and the groups of load real values;
the function acquisition module is used for acquiring a plurality of groups of prediction error normal distribution functions, and each group of prediction error normal distribution functions corresponds to the error of each group of load predicted values one by one;
the interval division module is used for carrying out interval division on the multiple groups of load predicted values according to the multiple groups of load predicted values and preset predicted interval boundary parameters and determining multiple predicted intervals;
the interval prediction model establishing module is used for generating a diversified load condition probability interval prediction model according to the multiple groups of prediction error normal distribution functions and the multiple prediction intervals;
the optimization processing module is used for performing iterative optimization on the diversified load condition probability interval prediction model, updating the boundary parameter of the prediction interval in the optimization process and determining the optimal boundary parameter value of the prediction interval; the optimal boundary parameter value is a prediction interval boundary parameter value which enables the range of each prediction interval to be minimum on the premise of ensuring that the real value of each group of loads falls in the corresponding prediction interval;
and the optimal prediction interval acquisition module is used for determining the optimal prediction interval of the diversified loads according to the boundary parameter of the optimal prediction interval.
7. The diversified load interval prediction device according to claim 6, wherein the function obtaining module comprises:
the mean standard deviation obtaining unit is used for determining an error mean value and an error standard deviation corresponding to the error of each group of load predicted values according to the error of each group of load predicted values;
and the function acquisition unit is used for determining a plurality of groups of prediction error normal distribution functions according to the error of each group of load prediction values, the error mean value and the error standard deviation.
8. The diversified load interval prediction device according to claim 6, wherein the interval division module comprises:
a normal distribution quantile obtaining unit, configured to determine, according to the preset prediction interval boundary parameter, a normal distribution quantile corresponding to the prediction interval boundary parameter;
the mean standard deviation obtaining unit is used for determining an error mean value and an error standard deviation corresponding to the error of each group of load predicted values according to the error of each group of load predicted values;
the interval upper and lower limit value acquisition unit is used for determining the upper and lower limit values of a group of prediction interval values according to the normal distribution quantile points, the error mean values and the error standard deviations aiming at any load prediction value;
a boundary value obtaining unit, configured to determine a boundary value of a group of predicted sections according to the predicted section boundary parameter and upper and lower limit values of the group of predicted sections;
a predicted block section obtaining unit configured to determine a predicted block section based on upper and lower limit values of the group of predicted block sections and a boundary value of the group of predicted block sections; a plurality of predicted block segments are obtained.
9. The diversified load interval prediction device according to claim 6, further comprising a prediction evaluation module for performing prediction evaluation on the optimal prediction interval after the optimal prediction interval of the diversified load is determined; the prediction evaluation is used for verifying the validity and correctness of the optimal prediction interval.
10. The diversified load interval prediction device according to claim 9, wherein the prediction interval evaluation comprises: average coverage error assessment, sensitivity assessment, and composite score value assessment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110211746.7A CN112734141B (en) | 2021-02-25 | 2021-02-25 | Diversified load interval prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110211746.7A CN112734141B (en) | 2021-02-25 | 2021-02-25 | Diversified load interval prediction method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112734141A true CN112734141A (en) | 2021-04-30 |
| CN112734141B CN112734141B (en) | 2023-01-20 |
Family
ID=75595405
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110211746.7A Active CN112734141B (en) | 2021-02-25 | 2021-02-25 | Diversified load interval prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112734141B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113642778A (en) * | 2021-07-21 | 2021-11-12 | 国网山东省电力公司经济技术研究院 | Power demand prediction framework considering multiple dimensions and multiple factors |
| CN113705867A (en) * | 2021-08-16 | 2021-11-26 | 浙江浙大中控信息技术有限公司 | Equipment abnormity diagnosis method based on energy consumption interval prediction |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109214605A (en) * | 2018-11-12 | 2019-01-15 | 国网山东省电力公司电力科学研究院 | Power-system short-term Load Probability prediction technique, apparatus and system |
| CN110263866A (en) * | 2019-06-24 | 2019-09-20 | 苏州智睿新能信息科技有限公司 | A kind of power consumer load setting prediction technique based on deep learning |
-
2021
- 2021-02-25 CN CN202110211746.7A patent/CN112734141B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109214605A (en) * | 2018-11-12 | 2019-01-15 | 国网山东省电力公司电力科学研究院 | Power-system short-term Load Probability prediction technique, apparatus and system |
| CN110263866A (en) * | 2019-06-24 | 2019-09-20 | 苏州智睿新能信息科技有限公司 | A kind of power consumer load setting prediction technique based on deep learning |
Non-Patent Citations (2)
| Title |
|---|
| 宗文婷等: "基于改进高斯过程回归模型的短期负荷区间预测", 《电力***及其自动化学报》 * |
| 黄南天等: "采用面积灰关联决策的高斯过程回归概率短期负荷预测", 《电力***自动化》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113642778A (en) * | 2021-07-21 | 2021-11-12 | 国网山东省电力公司经济技术研究院 | Power demand prediction framework considering multiple dimensions and multiple factors |
| CN113705867A (en) * | 2021-08-16 | 2021-11-26 | 浙江浙大中控信息技术有限公司 | Equipment abnormity diagnosis method based on energy consumption interval prediction |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112734141B (en) | 2023-01-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108375715B (en) | Power distribution network line fault risk day prediction method and system | |
| CN111628499B (en) | Method for evaluating new energy consumption capability of power distribution network considering multiple risk factors | |
| CN110826789B (en) | Power load prediction method and device based on power system and terminal equipment | |
| CN112734141B (en) | Diversified load interval prediction method and device | |
| CN111612275B (en) | Method and device for predicting load of regional user | |
| CN114595623A (en) | XGboost algorithm-based unit equipment reference value prediction method and system | |
| CN111639921A (en) | Intelligent equipment fault judgment and auxiliary disposal method based on expert system | |
| CN113408659A (en) | Building energy consumption integrated analysis method based on data mining | |
| CN115526258A (en) | Power system transient stability evaluation method based on Spearman correlation coefficient feature extraction | |
| CN111680712A (en) | Transformer oil temperature prediction method, device and system based on similar moments in the day | |
| CN117557009B (en) | Power efficiency monitoring method and system | |
| CN111080484A (en) | Method and device for monitoring abnormal data of power distribution network | |
| CN117595231B (en) | Intelligent power grid distribution management system and method thereof | |
| CN112418662A (en) | Power distribution network operation reliability analysis method using artificial neural network | |
| CN117494009A (en) | Electrical equipment state evaluation method based on insulating material pyrolysis analysis and cloud platform | |
| CN114926075B (en) | Machine part production scheduling method based on man-hour prediction | |
| CN113591266A (en) | Method and system for analyzing fault probability of electric energy meter | |
| Zafar et al. | A deep neural network-based optimal power flow approach for identifying network congestion and renewable energy generation curtailment | |
| CN108123436B (en) | Voltage out-of-limit prediction model based on principal component analysis and multiple regression algorithm | |
| CN112737120A (en) | Generation method and device of regional power grid control report and computer equipment | |
| CN112070303A (en) | Parameter-adaptive photovoltaic power ramp event hierarchical probabilistic prediction method | |
| CN114065460A (en) | Model processing method, storage medium and electronic device in thermal power generation system | |
| CN111967689A (en) | Model and method for wind power generation prediction by combining multivariate and stepwise linear regression and artificial neural network | |
| CN116739532B (en) | Energy project management method, system, terminal equipment and storage medium | |
| CN115498629B (en) | Scene enhanced power load operation control method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |