Selection of Building Blocks for Adaptive Grammatical Evolution in PSO

Abstract In this study, we used grammatical evolution to develop a customised particle swarm optimiser by incorporating adaptive building blocks. This makes the algorithm self-adaptable to the problem instance. Our objective is to provide the means to automatically generate novel population-based meta-heuristics by scoring the building blocks. We propose a new self-adapting algorithm by adaptive selection and scoring of the building blocks to solve multiple problem instances by reducing computation time and iteration count. To achieve our objective, we ranked building blocks that were extracted from a broad set of existing particle swarm optimisers and scored these during the evolutionary process. These scores were provided as an input to the evolutionary process that enabled the replacement of blocks of evolved solutions in cases where they were unable to improve the overall fitness. Our numerical experiments demonstrated that the proposed algorithm with adaptive building blocks reduced the iteration count and computation time with respect to PSO.