An adaptive Cuckoo search algorithm for optimisation

Cuckoo search is one of many nature-inspired algorithms used extensively to solve optimisation problems in different fields of engineering. It is a very effective in solving global optimisation because it is able to maintain balance between local and global random walks using switching parameter. The switching parameter for the original Cuckoo search algorithm is fixed at 25% and not enough studies have been done to assess the impact of dynamic switching parameter on the performance of Cuckoo search algorithm. This paper’s contribution is the development of three new Cuckoo search algorithms based on dynamically increasing switching parameters. The three new Cuckoo search algorithms are validated on ten mathematical test functions and their results compared to those of Cuckoo search algorithms with constant and dynamically decreasing switching parameters respectively. Finally, the simulations in this study indicate that, the Cuckoo search algorithm with exponentially increasing switching parameter outperformed the other Cuckoo search algorithms.