Nelder Mead Trained Neural Networks For Short Term Load Forecasting

This book proposes a new optimization algorithm for solving short term load forecasting problem. Globalized Nelder Mead is used for training of Artificial Neural Networks. Nelder Mead is fast optimization algorithm with no gradient calculation. The weights of Neural Networks are tuned with the help of Nelder Mead algorithm. To find proficiency of this algorithm, Australian Energy Market Operator (AEMO) data and California data are taken for testing. Results show that proposed algorithm outclasses other techniques in literature.

Autorentext

I, being Electrical Engineer, served in both industry and academic institutes. During my research period, I figured out that academic research is backbone of industry. If we stop finding solutions to problems or discovering new technologies, soon this world will go in dark ages. I had six (6) years of industrial, teaching and research experience.