A Hybrid Modern and Classical Algorithm for Indonesian Electricity Demand Forecasting
Department of Electrical Engineering
Universitas Negeri Gorontalo
In this paper, we would investigate a hybrid modern and classical algorithm (HMCA) in parameter optimization of electricity demand forecasting. Genetic algorithm (GA) has been successfully applied in optimization problems. As a modern algorithm, GA has a capability to explore the solutions in the global search area, but its drawback is the slow rate of convergence and high number of iterations. The Nelder-Mead is one of the classical algorithms using simplex search methods. This technique, when combining with a modern algorithm can be used for faster optimization processing. The test performance of the hybrid algorithm model (HAM) is conducted using data for Indonesian electricity demand. Results have shown that HMCA is better than GA in term of accuracy and number of iterations.
Key-Words: - Hybrid algorithm; Optimization; Genetic algorithm; Electricity demand forecasting