Probabilistic Load Forecasting Using an Improved Wavelet Neural Network Trained by Generalized Extreme Learning Machine

TY - JOUR

T1 - Probabilistic Load Forecasting Using an Improved Wavelet Neural Network Trained by Generalized Extreme Learning Machine

AU - Rafiei Foroushani, Mehdi

AU - Niknam, Taher

AU - Aghaei, Jamshid

AU - Shafie-khah, Miadreza

AU - P.S.Catalão, João

PY - 2018/11

Y1 - 2018/11

N2 - Competitive transactions resulting from recent restructuring of the electricity market, have made achieving a precise and reliable load forecasting, especially probabilistic load forecasting, an important topic. Hence, this paper presents a novel hybrid method of probabilistic electricity load forecasting, including generalized extreme learning machine for training an improved wavelet neural network, wavelet preprocessing and bootstrapping. In the proposed method, the forecasting model and data noise uncertainties are taken into account while the output of the model is the load probabilistic interval. In order to validate the method, it is implemented on the Ontario and Australian electricity markets data. Also, in order to remove the influence of model parameters and data on performance validation, Friedman and post-hoc tests, which are non-parametric tests, are applied to the proposed method. The results demonstrate the high performance, accuracy, and reliability of the proposed method.

AB - Competitive transactions resulting from recent restructuring of the electricity market, have made achieving a precise and reliable load forecasting, especially probabilistic load forecasting, an important topic. Hence, this paper presents a novel hybrid method of probabilistic electricity load forecasting, including generalized extreme learning machine for training an improved wavelet neural network, wavelet preprocessing and bootstrapping. In the proposed method, the forecasting model and data noise uncertainties are taken into account while the output of the model is the load probabilistic interval. In order to validate the method, it is implemented on the Ontario and Australian electricity markets data. Also, in order to remove the influence of model parameters and data on performance validation, Friedman and post-hoc tests, which are non-parametric tests, are applied to the proposed method. The results demonstrate the high performance, accuracy, and reliability of the proposed method.

KW - Probabilistic forecasting

KW - bootstrapping

KW - generalized extreme learning machine

KW - improved wavelet neural network

KW - wavelet processing

UR - http://www.scopus.com/inward/record.url?scp=85042362642&partnerID=8YFLogxK

U2 - 10.1109/TSG.2018.2807845

DO - 10.1109/TSG.2018.2807845

M3 - Journal article

SN - 1949-3053

VL - 9

SP - 6961

EP - 6971

JO - IEEE Transactions on Smart Grid

JF - IEEE Transactions on Smart Grid

IS - 6

M1 - 8298533

ER -