Novel metaheuristic approach: integration of variables method (IVM) and human-machine interaction for subjective evaluation

Abstract

Metaheuristics is recognized as the most practical approach in simulation based global optimization and during investigation of the state-of-the-art optimization approaches from local searches over evolutionary computation methods to estimation of distribution algorithms. In this study a new evolutionary metaheuristic approach “The Integration of Variables Method (IVM)” is devised for global optimization, that methodology is characterized by making vide solutions codes in an objective population aided by any composition of operators. The Genetic Algorithms (GA) are incorporated in this methodology; where genetic operators are used to evolve populations classifiers in creation of new concrete heuristics solutions for decision making tasks. In this work, the main consideration is devoted to the exploration of Extremes value distribution of Functions with multi Variables Codes in brain decision making process and Brain Computer Interface (BCI) tasks. The properties of Extremes value distribution algorithm are characterized by the diversity of measurements in populations of available solutions. Additionally, while generating adequate solutions of the basic objectives, subjective indicators of human-machine interaction are used to differentiate the characteristics of the different solutions in a population. The results obtained from this metaheuristic approach are compared with results of Genetic Algorithms (GA) in case study related with Optimal Multiple Objective Design and Progressive Cutting Dies implemented in a CAD system.