A New Hybrid ARAR and Neural Network Model for Multi-Step Ahead Wind Speed Forecasting in Three Regions of Pakistan

Wind is one of the most essential sources of clean, environmental friendly, socially constructive, economically beneficial, and renewable energy. To intuit the potential of this energy in a region the accurate wind speed modeling and forecasting are crucially important, even for planning, conversion of wind energy to electricity, energy trading, and reducing instability. However, accurate prediction is difficult due to intermittency and intrinsic complexity in wind speed data. This study aims to suggest a more appropriate model for accurate wind speed forecasting in the Jhimpir, Gharo, and Talhar, regions of Sindh, Pakistan. Therefore, the present study combined the Autoregressive-Autoregressive (ARAR) and Artificial Neural Network (ANN) models to propose a new hybrid ARAR-ANN model for better prediction by precisely capturing different patterns of the wind speed time-series data sets. The proposed hybrid model is efficient in modeling, reducing statistical errors, and forecasting the wind speed effectively. The performance of the proposed hybrid ARAR-ANN model is compared using three error-statistics and Nash-Sutcliffe efficiency-coefficient. The empirical results of the four performance indices fully demonstrated the superiority of the hybrid ARAR-ANN model than persistence model, ARAR, ANN and SVM. Indeed, the proposed model is an effective and feasible approach for wind speed forecasting.