A Hybrid bVAR-NARX Wind Power Forecasting Model Based on Wind and Load Demand Correlation: A Case Study of ERCOT’s System from an ISO’s Perspective

Abstract

Wind power forecasting tools are as crucial for independent system operators (ISO) as it is for individual wind farm owners. ISO integrate wind power forecasting tools into power system operations for a number of applications and processes such as balancing energy, regulation and energy reserve services. To deal with the challenges that have surfaced in implementing wind power forecasting tools from an ISO’s perspective, a hybrid model is proposed which is based on the correlation between wind power and load demand. The proposed model has used a bi-sectional vector auto-regressive model in conjunction with a nonlinear auto-regressive with exogenous output model forming a hybrid model that gives better results on real-world data sets as well. The case study taken for validating the model is the Electric Reliability Council of Texas system that has seen tremendous increase in the installed capacity of wind power generation sources in the recent past.

Keywords:

• wind power forecasting
• wind and load correlation
• multivariate modeling
• neural network (NARX)
• vector autoregressive model
• hybrid model
• ERCOT wind data set
• descriptive analysis
• objective analysis
• correlation functions

Acknowledgments

The authors express their sincere gratitude towards Mr. Heistrene Konstantine and Mrs. Laly Heistrene for their moral support, constant encouragement and guidance all throughout the research work. We would also like to thank Mr. Ravi Pathak, co-founder of Tatvic, Inc, for his insights into the world of forecasting and Mr. Parth Upadhyay for his relentless writing assistance.

Additional information

Notes on contributors

Leena Heistrene

Leena Heistrene graduated in 2002 and worked as an engineer with Torrent Power Ltd., India, from 2002 to 2008. She completed her post-graduation in 2012 and joined the field of academics. She is currently working as a lecturer in Electrical Engineering Department of Pandit Deendayal Petroleum University, India. Her main research interests are wind/solar/load forecasting and power system optimization, especially market clearing strategies in the wake of high penetration of renewable sources of energy.

Poonam Mishra

Poonam Mishra received her Ph. D. degree in the year 2010 in applied mathematics, since then she is associated with Mathematics department of Pandit Deendayal Petroleum University, India. She has 30 publications on her name in various peer-reviewed, national and international journals. She has many chapters in edited books to her credit and has presented several papers at international conferences outside India. Her core research interest is mathematical modelling of real problems, optimization and supply chain management.

Makarand Lokhande

Makarand Lokhande (IEEE member) received the B.E Degree in Electrical Engineering from Nagpur University in 2001, the M.E. Degree in Power System from the University of Pune in 2003 and Ph.D. Degree from Indian Institute of Technology Bombay in 2010. He has been working as an Assistant Professor in the Department of Electrical Engineering of Visvesvaraya National Institute of Technology, India since 2015. Prior to that he was also associated with Sardar Vallabhbhai National Institute of Technology, Surat, Gujarat, India and Pandit Deendayal Petroleum University Gujarat, India as an Assistant Professor. His research activities are related to power electronics, machines, electric vehicles and photo-voltaic systems.