Analysis of evolutionary operators for ICHEA in solving constraint optimization problems

Intelligent constraint handling evolutionary algorithm (ICHEA) is a recently proposed variation of evolutionary algorithm (EA) that solves real-valued constraint satisfaction problems (CSPs) efficiently. Initially it was designed to solve CSPs only, however, it has been shown effective in solving static and dynamic constraint optimization problems as well (Sharma and Sharma, 2012). ICHEA has ability to extract and exploit information from constraints that guides its evolutionary search operators in contrast to traditional EAs that are `blind' to constraints. Several variations of EAs have been proposed to solve constraint/optimization problems in the literature. Many articles have the main objective to show the efficiency of one algorithm by outperforming other algorithms in terms of fewer evaluations, solutions closer to the global known solutions or one that takes less processing time. There are not many articles that provide a systematic model on examining the multiple operators of an algorithm to evaluate their efficiency or effectiveness in a given environment. An algorithm with multiple operators like ICHEA generally gives mediocre results when only a single operator is applied in the algorithm, however, collectively with other operator(s) good solutions are obtained. In this paper we describe the enhanced ICHEA with additional operators that produce better results on benchmark problems than previously published in (Sharma and Sharma, 2012). It describes how the operators behave in search for the optimal solution and impact the environment in terms of population diversity, improvement in solutions and genetic drift.