Output details
12 - Aeronautical, Mechanical, Chemical and Manufacturing Engineering
University of Birmingham : A - Mechanical Engineering
Benchmarking and comparison of nature-inspired population-based continuous optimisation algorithms
This paper describes an investigation into four nature-inspired continuous optimisation methods: the authors’ Bees Algorithm and Evolutionary Algorithms, Particle Swarm Optimisation, and Artificial Bee Colony. The aim was to understand and compare the specific capabilities of the algorithms using twenty-five new minimisation benchmarks designed by the authors. The results highlight the strengths and weaknesses of the algorithms. The existence and extent of origin and alignment biases related to the use of different recombination operators were uncovered for the first time. The work reveals interesting regularities that will help to guide the choice and configuration of optimisation algorithms for practical applications.