A comparative evaluation for identifying the suitability of extreme learning machine to predict horizontal global solar radiation

摘要: Abstract In this paper, the extreme learning machine (ELM) is employed to predict horizontal global solar radiation (HGSR). For purpose, capability of developed ELM method appraised statistically for prediction monthly mean daily HGSR using three different types input parameters: (1) sunshine duration-based (SDB), (2) difference temperature-based (TB) and (3) multiple parameters-based (MPB). The long-term measured data sets collected city Shiraz situated in Fars province Iran have been utilized as a case study. predicted via compared with those support vector (SVM), genetic programming (GP) artificial neural network (ANN) ensure precision ELM. It found that higher accuracy can be obtained by estimation all techniques. computational results prove highly accurate reliable shows performance than SVM, GP ANN. model, absolute percentage error, bias root square relative error coefficient determination are 2.2518%, 0.4343 MJ/m 2 , 0.5882 MJ/m 2.9757% 0.9865, respectively. By conducting further verification, it also offers high superiority over four empirical models established study an intelligent model from literature. final analysis, proper sensitivity analysis performed identify influence considered elements on which reveal significance appropriate selection parameters boost algorithm. nutshell, comparative clearly specify technique provide predictions existing