CN113850443A - Short-term power load interval prediction method based on nonparametric Bootstrap error sampling - Google Patents
Short-term power load interval prediction method based on nonparametric Bootstrap error sampling Download PDFInfo
- Publication number
- CN113850443A CN113850443A CN202111197375.8A CN202111197375A CN113850443A CN 113850443 A CN113850443 A CN 113850443A CN 202111197375 A CN202111197375 A CN 202111197375A CN 113850443 A CN113850443 A CN 113850443A
- Authority
- CN
- China
- Prior art keywords
- interval
- prediction
- error
- power load
- short
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000005070 sampling Methods 0.000 title claims abstract description 17
- 238000012549 training Methods 0.000 claims abstract description 7
- 238000007781 pre-processing Methods 0.000 claims abstract description 4
- 238000012360 testing method Methods 0.000 claims description 5
- 239000011159 matrix material Substances 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 abstract description 2
- 230000000694 effects Effects 0.000 description 3
- 238000007726 management method Methods 0.000 description 3
- 241001123248 Arma Species 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- YHXISWVBGDMDLQ-UHFFFAOYSA-N moclobemide Chemical compound C1=CC(Cl)=CC=C1C(=O)NCCN1CCOCC1 YHXISWVBGDMDLQ-UHFFFAOYSA-N 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012300 Sequence Analysis Methods 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000001932 seasonal effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012731 temporal analysis Methods 0.000 description 1
- 238000000700 time series analysis 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- 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
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Entrepreneurship & Innovation (AREA)
- Theoretical Computer Science (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a short-term power load interval prediction method based on nonparametric Bootstrap error sampling, and belongs to the technical field of short-term power load prediction. The method comprises the steps of preprocessing original time sequence data by adopting a sliding window method, wherein the sliding window method is a technology aiming at real-time updating of the time sequence data and aims to construct a characteristic data set for model training; establishing an ELM-AdaBoost model to fit and determine the load predicted value of the point; carrying out interval division on the actual load value, and sampling the error in each interval by adopting a nonparametric Bootstrap method to obtain a confidence interval of a predicted error; establishing related statistics, and converting the confidence interval of the error into a prediction interval of the load value; the model reduces the uncertainty of load prediction and the prediction error thereof, and has certain applicability.
Description
Technical Field
The invention belongs to the technical field of short-term power load prediction, and relates to a short-term power load interval prediction method based on nonparametric Bootstrap error sampling.
Background
With the improvement of the living standard of people, the acceleration of industrialization and the growth of population, the demand of electric power is also continuously increased. This inevitably increases the complexity and uncertainty of the operation of the power system. However, stable operation of the grid is based on supply-demand balancing on all time scales. Therefore, accurate demand forecasting has become an indispensable task in power system management. Accurate estimation of future power load changes is of great significance to decision making and operation of the power grid, as well as to power companies and users. Accurate load prediction not only provides economic, reliable and sustainable electric energy for society and people, but also provides effective decision basis for investment planning of electric power markets. Short-term power load forecasts are typically power load forecasts over the range of hours or weeks. It is an important task in the management of power systems, and relates to maintenance planning, operation and planning of power systems. Therefore, accurate and stable short-term power load prediction can greatly reduce the operation cost of a power grid company, and help and remind a dispatcher to ensure the safe operation of a power system.
Due to the volatility and instability of renewable energy sources in power structures, their impact on grid stability is becoming more and more significant. The trend of future power demand is mastered, and the realization of power supply and demand balance is more and more difficult. Various new researches have been carried out in the field of power load prediction, and a large number of short-term power load prediction models exist at present. The models are mainly divided into two categories of traditional prediction methods and current prediction technologies, and are respectively provided based on statistical modeling and artificial intelligence.
Traditional statistical prediction methods include time series analysis and regression analysis, and both ARMA and ARIMA methods require conversion of raw data into stationary time series. Furthermore, the ARMA and ARIMA methods only deal with linear sequence analysis and are not applicable to non-linear sequences in power load data. With the mature application of the SVR model in load prediction, the SVR model can solve the complex nonlinear regression problem, has advantages on the small sample problem, but is difficult to adapt to the large-scale sample. With the development of artificial intelligence, machine learning methods have emerged, mainly based on neural network prediction techniques. However, the machine learning method has some disadvantages that it is difficult to process missing data and it is easy to have overfitting problem. Both traditional predictive models and modern neural network predictive models produce deterministic point predictions, one value at a time. Since the load has uncertainty and volatility due to the influence of climate change, seasonal factors and the like, the point prediction method cannot provide good prediction performance, and a stable power load prediction model needs to be established. The power load data is essentially a random, non-linear sequence that contains many uncertain influencing factors. Therefore, probabilistic predictions or interval predictions are needed to provide more information about the likely deviation of the prediction.
Disclosure of Invention
In view of the above, the present invention provides a short-term power load interval prediction method based on nonparametric Bootstrap error sampling, which samples prediction errors in different intervals, and converts interval prediction statistics into interval prediction, thereby implementing an interval prediction structure and implementing accurate prediction of a short-term power load.
In order to achieve the purpose, the invention provides the following technical scheme:
a short-term power load interval prediction method based on nonparametric Bootstrap error sampling specifically comprises the following steps:
s1: preprocessing the acquired original load time sequence data by adopting a sliding window method, establishing a short-term power load data set, and carrying out data processing on the data set according to the following steps of 7: 3, dividing the ratio into a training set and a testing set;
s2: establishing an ELM-AdaBoost model, inputting the preprocessed data set, and obtaining a predicted value of a determined point;
s3: dividing the test set into different sections according to the determined interval, calculating the prediction error in each section, and merging adjacent sections with smaller data volume;
s4: bootstrap sampling is carried out on the error data in each interval to obtain confidence intervals of the errors in different interval sections under different confidence degrees;
s5: and determining the section to which the observation value belongs, constructing statistics related to the prediction error, and converting the confidence interval of the error into a prediction interval of a prediction value, thereby realizing interval prediction of the power load.
Further, in step S1, the collected original load time-series data is preprocessed by using a sliding window method, and the time-series data is converted into matrix data to create a short-term power load data set.
Further, in step S2, the established ELM-AdaBoost model is a regression model based on an Extreme Learning Machine (ELM), T weak predictors are obtained through training of the AdaBoost algorithm, corresponding weights are given to the T weak predictors, and a strong predictor is combined to form a strong predictor
Wherein alpha istIs the weight of the weak predictor, ht(X) is weak predictor, T is 1, …, T.
Further, in step S4, the confidence interval EI of the error in the different segment is equal to (EI)1,EI2,…,EIi,…,EIn) Wherein
n is the number of the intervals,
respectively, the lower limit and the upper limit of the i-th interval error.
The invention has the beneficial effects that:
(1) the invention adopts a sliding window method to preprocess time series data so as to ensure that the data can adapt to the normal training of the model.
(2) The invention adopts a nonparametric Bootstrap sampling method to carry out statistical analysis on the prediction error sequence, constructs interval prediction and improves the prediction precision.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention may be realized and attained by the means of the instrumentalities and combinations particularly pointed out hereinafter.
Drawings
For the purposes of promoting a better understanding of the objects, aspects and advantages of the invention, reference will now be made to the following detailed description taken in conjunction with the accompanying drawings in which:
FIG. 1 is a flow chart of a short-term power load interval prediction method based on nonparametric Bootstrap error sampling according to the present invention;
fig. 2 shows the predicted result of the short-term power load interval in queensland, australia in this example.
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention in a schematic way, and the features in the following embodiments and examples may be combined with each other without conflict.
Referring to fig. 1, the short-term power load interval prediction method based on nonparametric Bootstrap error sampling according to the invention preprocesses the acquired original load time sequence data, converts the time sequence data into matrix data through a sliding window method, and establishes a short-term power load data set; performing point prediction on the power load data by using an ELM-AdaBoost model to obtain a point prediction value at each moment; sampling errors in different intervals by a Bootstrap error sampling method, establishing an interval prediction method to perform interval prediction on a load curve of three days in the future, and specifically comprising the following steps:
step 1: when preprocessing the acquired data, the window width is slid to 4, and as shown in fig. 2, the time-series data x is updated to (x)1,x2,x3,…,xn-1,xn) Conversion into matrix data (X, Y); x is an input and Y is an output.
Wherein the content of the first and second substances,
Y＝[Y1,Y2,…,YN-1,YN]＝[x5,x6,…,xn-1,xn]。
step 2: an ELM-based regression model is used, T weak predictors are obtained through AdaBoost algorithm training, corresponding weights are given to the T weak predictors, and a strong predictor is combined
Wherein alpha istIs the weight of the weak predictor, ht(X) is weak predictor, T is 1, …, T. Namely the ELM-AdaBoost model. Load data of three days in the future is predicted by adopting the model, and a predicted value of a determined point is obtained;
and step 3: dividing the test set at equal intervals, calculating the prediction error in each interval, and merging the adjacent intervals with smaller data volume;
and 4, step 4: bootstrap sampling is carried out on the error data in each interval, and confidence intervals EI (EI) of the errors in different interval are obtained under different confidence degrees1,EI2,…,EIi,…,EIn) Wherein
n is the number of the intervals,
the lower limit and the upper limit of the ith interval error respectively;
and 5: and determining the section to which the observation value belongs, constructing statistics related to the prediction error, and converting the confidence interval of the error into a prediction interval of a prediction value, thereby realizing interval prediction of the power load.
Table 1 table of the result of the final interval division in this embodiment
TABLE 2 comparison of the implementation of the prediction model of the present invention with the existing prediction models
As can be seen from Table 2, the prediction effect of the ELM-AdaBoost model is the best, the average value of FICP is 66.32%, which is 1.16%, 2.89% and 4.05% higher than that of BP model, ELM model and BP-AdaBoost model respectively; MWFI average value is 1.07%; the variation is not very large with respect to the other several models. The prediction effect of the BP-AdaBoost model is the worst, and the average values of FICP and MWFI are respectively 62.27% and 1.10%.
The results show that the ELM-AdaBoost is used as a point prediction model and the interval prediction method of nonparametric Bootstrap sampling can provide more accurate prediction results for the power system, so that not only is the power management more convenient, but also the operation arrangement of the power grid is more reasonable. An ELM-AdaBoost model is adopted as a point prediction model, and interval prediction results under different confidence levels are shown in FIG. 2.
Finally, the above embodiments are only intended to illustrate the technical solutions of the present invention and not to limit the present invention, and although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions, and all of them should be covered by the claims of the present invention.
Claims (4)
1. A short-term power load interval prediction method based on nonparametric Bootstrap error sampling is characterized by comprising the following steps:
s1: preprocessing the acquired original load time sequence data, establishing a short-term power load data set, and dividing the data set into a training set and a testing set according to a proportion;
s2: establishing an ELM-AdaBoost model, inputting the preprocessed data set, and obtaining a predicted value of a determined point;
s3: dividing the test set into different sections according to the determined interval, calculating the prediction error in each section, and merging adjacent sections with smaller data volume;
s4: bootstrap sampling is carried out on the error data in each interval to obtain confidence intervals of the errors in different interval sections under different confidence degrees;
s5: and determining the section to which the observation value belongs, constructing statistics related to the prediction error, and converting the confidence interval of the error into a prediction interval of a prediction value, thereby realizing interval prediction of the power load.
2. The method for predicting a short-term power load interval according to claim 1, wherein in step S1, the collected time-series data of the original load is preprocessed by a sliding window method, and the time-series data is converted into matrix data to create a short-term power load data set.
3. The short-term power load interval prediction method according to claim 1, wherein in step S2, the established ELM-AdaBoost model is a regression model based on ELM, T weak predictors are obtained through training of AdaBoost algorithm, corresponding weights are given to the T weak predictors, and a strong predictor is combined by combining the T weak predictors with the corresponding weights
Wherein alpha istIs the weight of the weak predictor, ht(X) is weak predictor, T is 1, …, T.
4. The method as claimed in claim 1, wherein in step S4, the confidence interval EI of the error in different segments is (EI)1,EI2,…,EIi,…,EIn) Wherein
n is the number of the intervals,
respectively, the lower limit and the upper limit of the i-th interval error.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111197375.8A CN113850443A (en) | 2021-10-14 | 2021-10-14 | Short-term power load interval prediction method based on nonparametric Bootstrap error sampling |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111197375.8A CN113850443A (en) | 2021-10-14 | 2021-10-14 | Short-term power load interval prediction method based on nonparametric Bootstrap error sampling |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113850443A true CN113850443A (en) | 2021-12-28 |
Family
ID=78978402
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111197375.8A Pending CN113850443A (en) | 2021-10-14 | 2021-10-14 | Short-term power load interval prediction method based on nonparametric Bootstrap error sampling |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113850443A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117408299A (en) * | 2023-09-08 | 2024-01-16 | 国网湖北省电力有限公司宜昌供电公司 | Deep learning-based prediction method for concentration of dissolved gas in transformer oil |
-
2021
- 2021-10-14 CN CN202111197375.8A patent/CN113850443A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117408299A (en) * | 2023-09-08 | 2024-01-16 | 国网湖北省电力有限公司宜昌供电公司 | Deep learning-based prediction method for concentration of dissolved gas in transformer oil |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113962364B (en) | Multi-factor power load prediction method based on deep learning | |
| CN109711620B (en) | Short-term power load prediction method based on GRU neural network and transfer learning | |
| CN111260136A (en) | Building short-term load prediction method based on ARIMA-LSTM combined model | |
| CN111080032A (en) | Load prediction method based on Transformer structure | |
| CN113554466B (en) | Short-term electricity consumption prediction model construction method, prediction method and device | |
| CN111695724B (en) | Wind speed prediction method based on hybrid neural network model | |
| CN112329990A (en) | User power load prediction method based on LSTM-BP neural network | |
| CN113344288B (en) | Cascade hydropower station group water level prediction method and device and computer readable storage medium | |
| CN112149890A (en) | Comprehensive energy load prediction method and system based on user energy label | |
| CN110334847A (en) | Based on the wind power prediction method for improving K-means cluster and support vector machines | |
| CN115860177A (en) | Photovoltaic power generation power prediction method based on combined machine learning model and application thereof | |
| CN112990587A (en) | Method, system, equipment and medium for accurately predicting power consumption of transformer area | |
| CN114021848A (en) | Generating capacity demand prediction method based on LSTM deep learning | |
| Zhang et al. | Interval prediction of ultra-short-term photovoltaic power based on a hybrid model | |
| CN115545333A (en) | Method for predicting load curve of multi-load daily-type power distribution network | |
| CN113850443A (en) | Short-term power load interval prediction method based on nonparametric Bootstrap error sampling | |
| CN110852492A (en) | Photovoltaic power ultra-short-term prediction method for finding similarity based on Mahalanobis distance | |
| CN114611757A (en) | Electric power system short-term load prediction method based on genetic algorithm and improved depth residual error network | |
| CN114595542A (en) | Line loss prediction method | |
| CN112052985B (en) | Middle-short-term low-voltage prediction method based on lightgbm | |
| CN113139698A (en) | Load prediction method, device and equipment | |
| CN116826710A (en) | Peak clipping strategy recommendation method and device based on load prediction and storage medium | |
| CN116231749A (en) | New energy power system dispatching method based on digital twin | |
| CN116307111A (en) | Reactive load prediction method based on K-means clustering and random forest algorithm | |
| CN115759343A (en) | E-LSTM-based user electric quantity prediction 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 |