
Short‐term load forecast of electrical power system by radial basis function neural network and new stochastic search algorithm
Author(s) -
Abedinia Oveis,
Amjady Nima
Publication year - 2016
Publication title -
international transactions on electrical energy systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.428
H-Index - 42
ISSN - 2050-7038
DOI - 10.1002/etep.2160
Subject(s) - overfitting , artificial neural network , radial basis function , computer science , term (time) , electric power system , mathematical optimization , function (biology) , artificial intelligence , algorithm , power (physics) , mathematics , physics , quantum mechanics , evolutionary biology , biology
Summary In this paper a new model of radial basis function (RBF) neural network based on a novel stochastic search algorithm is presented for short‐term load forecast (STLF). STLF is an important operation function in both regulated and deregulated power systems. Accurate STLF is effective for area generation control and resource dispatch as well as electricity market clearing. The proposed STLF method optimizes the structure of the RBF‐based forecasting engine. Random selection of the number of hidden neurons may cause overfitting or underfitting problem of the network. For this purpose, a new stochastic search algorithm is presented to find the optimum number of neurons for the hidden layer. To demonstrate the effectiveness of the proposed STLF approach, it is tested on three real‐world engineering case studies, and the obtained results are compared with the results of several other recently published STLF methods. These comparisons confirm the validity of the developed approach. Copyright © 2015 John Wiley & Sons, Ltd.