A variant of environmental adaptation method with real parameter encoding and its application in economic load dispatch problem

Abstract

Environmental Adaptation Method (EAM) and Improved Environmental Adaptation Method (IEAM) were proposed to solve optimization problems with the biological theory of adaptation in mind. Both of these algorithms work with binary encoding, and their performance is comparable with other state-of-art algorithms. To further improve the performance of these algorithms, some major changes are incorporated into the proposed algorithm. The proposed algorithm works with the real value parameter encoding, and, in order to maintain significant convergence rate and diversity, it maintains a balance between exploitation and exploration. The choice to explore or exploit a solution depends on the fitness of the individual. The performance of the proposed algorithm is compared with 17 state-of-art algorithms in 2-D, 3-D, 5-D, 10-D and 20-D dimensions using the COCO (COmparing Continuous Optimisers) framework with Black-Box Optimization Benchmarking (BBOB) functions. It outperforms all other algorithms in 3-D and 5-D, and its performance is comparable to other algorithms for other dimensions. In addition, IEAM-R has been applied to the real world problem of economic load dispatch, and its results demonstrate that it gives minimum fuel cost when compared to other algorithms in different cases.