Article Preview
TopIntroduction
The economic load dispatch (ELD) problem is one of the mathematical optimization issues in operation of the modern power system to schedule the committed generating unit outputs so as to meet the essential load demand while satisfying all equality and inequality constraints of the system at minimum operating cost (Cai, Ma, Li, Peng, 2007; Baskar & Mohan, 2008; Coelho & Lee, 2008). To get a better solution, attract many researchers’ attention because of good solutions from the ELD problem would result in great economic benefits. Several optimization methods have been applied to solve the ELD problems in the past decades. The most popular predictable approach to solve ELD problem is lambda-iteration method (Wood & Wollenberg, 1996; Chen & Wang, 1993) where continuous and monotonically increasing cost function is needed. Some other conventional approaches have been used to solve the ELD problems such as gradient method (Dodu, Martin, Merlin, & Pouget, 1972), lagrangian relaxation algorithm (El-Keib, Ma, & Hart, 1994), quadratic programming (Dos, Coelho, & Mariani, 2006)”, linear programming algorithm (Jabr, Coonick, & Cory, 2000), and etc. Also, many artificial intelligence technologies have been successfully used to solve the ELD problem more efficiently. The heuristic methods are genetic algorithm (Nanda & Badri, 2002; Lee, Lin, Liao, & Tsao, 2011),evolutionary programming (EP) (Yang, Yang, & Huang, 1996; Sinha, Chakrabarti, & Chattopadhyay, 2013), particle swarm optimization (Park, Lee, Shin, Lee, 2005), chaotic particle swarm optimization (Cai, Ma, Li, & Peng, 2007; Cai, Ma, Li, Peng, 2009), hybrid DE (Wang, Chiou, & Liu, 2007), ant colony optimization (ACO) (Yu, & Song, 2001), and chaotic ant swarm optimization (Cai, Ma, Li, Li, & Peng, 2010) have been proposed to solve different types of ELD problems. The gravitational search algorithm (GSA) (Roy, Paul, & Sultana, 2014; Roy, Mandal, & Bhattacharya, 2012; Bhattacharya & Roy, 2012), is a new heuristic algorithms inspired by the Newtonian laws of gravity and motion. Biogeography based optimization (BBO) (Simon, 2008) is one of the latest optimization algorithms. It describes how a natural process can be modeled to solve general optimization problems and has been applied to optimization ELD problems in (Roy, Ghoshal, & Thakur, 2010; Roy & Mandal, 2011) and krill herd algorithm (KHA) (Mandal, Roy, & Mandal, (2014) have been presented to solve ELD. The objective function of the most popular existing approaches should be continuous and differentiable but, the cost function considering valve-point effects are not differentiable. Some heuristic methodologies suffer from premature convergence, especially for ELD problem with multiple local optimums. Due to mentioned drawbacks, solutions of ELD lead to a non-optimal solution with a great economic loss.