Hybrid ensemble intelligent model based on wavelet transform, swarm intelligence and artificial neural network for electricity demand forecasting. (March 2021)

• Record Type: Journal Article
• Authors: Ofori-Ntow Jnr, Eric; Ziggah, Yao Yevenyo; Relvas, Susana
• Is Part Of: Sustainable cities and society. Volume 66(2021)
• Publication Date: 2021-03
• Subjects: AFD adaptive fourier decomposition -- AI artificial intelligent -- ANFIS adaptive neuro-fuzzy inference system -- ANN artificial neural network -- ARIMA autoregressive integrated moving average -- BPNN backpropagation neural network -- CWT continuous wavelet transform -- DBN deep belief network -- DWT discrete wavelet transform -- FANGM fractional order accumulation nonlinear grey model -- GA general algorithm -- GM grey model -- KF kalman filtering -- MGM metabolic grey model -- MLP multi-layer perceptron -- NMGM nonlinear metabolic grey model -- OMGM optimized metabolic grey model -- ONMGM optimized nonlinear metabolic grey model -- PSO particle swarm optimization -- RBFNN radial basis function neural network -- SAPSO self-adaptive particle swarm optimization -- SARIMA sinusoidal regressive integrated moving average -- SCASVM sine cosine algorithm and support vector machine -- SVM support vector machine -- VMD variational mode decomposition -- WNN wavelet neural network; Hybrid short-term load forecasting -- Discrete wavelet transform -- Optimization -- Ensemble method; Sustainable urban development -- Periodicals; Sustainable buildings -- Periodicals; Urban ecology (Sociology) -- Periodicals; 307.76
• DOI: 10.1016/j.scs.2020.102679
• Languages: English
• ISSNs: 2210-6707

Abstract

Highlights: A hybrid ensemble intelligent model based on wavelet transform, swarm intelligence and artificial neural network is developed. The model is capable of handling high volatility and nonlinear behaviour of time seies data. The proposed DWT-PSO-RBFNN model produces forecasting accuracy of 97.5778 %. The proposed model has a lower average relative error in comparison to other models. Abstract: Availability of electrical energy affects many facets of an entire economy of a country. This has made short-term electrical load forecasting an important area in recent years for policy makers and academic researchers. However, it has been found that the actual load series exhibit some complex behaviours which are often characterised by nonlinearity, nonstationarity, and temporal variations. In this study, a three-level hybrid ensemble short-term load forecasting method consisting of Discrete Wavelet Transform (DWT), Particle Swarm Optimization (PSO), and Radial Basis Function Neural Network (RBFNN) is proposed. The DWT is applied to decompose the data to get a well-behaved requisite series for forecasting since the data becomes stable before using PSO. PSO is used to obtain the required optimal adjustable parameters of the RBFNN for the forecasting. The proposed hybrid ensemble method (DWT-PSO-RBFNN) was evaluated using Ghana Grid Company daily average demand data from 1 st December 2018 to 30th November 2019. The DWT-PSO-RBFNN approach was compared with three other DWT coupling methods namely RBFNN, Backpropagation Neural Network (BPNN), and Self Adaptive Differential Evolution – Extreme Learning Machine (SaDE-ELM). The statistical analysis revealed that the proposed method performed better based on MAPE, MAD, and RMSE emphasizing its great potential.