An adaptive hybrid fractal model for short-term load forecasting in power systems

•Short-term load forecasting (STLF) of power loads involves fractal features, due to complexity of power systems.•Composite linear fractal interpolation function (CLFIF) modeling better captures irregularities and fluctuations of power loads.•Iterative learning (IL) and chimp optimization algorithm...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Electric power systems research Ročník 207; s. 107858
Hlavní autori: Li, Xiaolan, Zhou, Jun
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Amsterdam Elsevier B.V 01.06.2022
Elsevier Science Ltd
Predmet:
Adaptive
Algorithms
Chimp optimization algorithm
Composite linear fractal interpolation function
Electricity distribution
Forecasting
Fractal models
Fractals
Interpolation
Iterative learning
Iterative methods
Machine learning
Model accuracy
Numerical analysis
Optimization
Optimization algorithms
Parameterization
Short-term load forecasting
Time series
Chimp optimization algorithm
Adaptive
Composite linear fractal interpolation function
Short-term load forecasting
Iterative learning
ISSN:0378-7796, 1873-2046
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:•Short-term load forecasting (STLF) of power loads involves fractal features, due to complexity of power systems.•Composite linear fractal interpolation function (CLFIF) modeling better captures irregularities and fluctuations of power loads.•Iterative learning (IL) and chimp optimization algorithm (ChOA) are used for adaptive parametrization of CLFIF.•Hybrid CLFIF-IL-ChOA modeling is proposed for STLF, which obtains high accuracy without further relevant data.•Comparisons in between CLFIF-IL-ChOA and others are included; effectiveness of CLFIF-IL-ChOA is verified. This paper develops a hybrid short-term load forecasting (STLF) modeling with adaptive parametrization, based on composite linear fractal interpolation function (CLFIF), iterative learning (IL) and chimp optimization algorithm (ChOA). More precisely, after selecting similar days in power load data, an amendatory CLFIF model is constructed for an-hour-ahead prediction in terms of hourly load curves. Then, iterative learning based on ChOA optimizes the parameters of the amendatory CLFIF model with higher accuracy. Moreover, to confirm effectiveness of the CLFIF-IL-ChOA model, numerical examples are tested on the historical power load data from PJM and ENTSOE. The numerical results show that the proposed method can obtain higher accuracy, compared to some common methods about time series analysis.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0378-7796
1873-2046
DOI:10.1016/j.epsr.2022.107858