An Improved dynamic self-adaption cuckoo search algorithm based on collaboration between subpopulations

In order to improve convergence rate and optimization precision of the cuckoo search (CS) algorithm, an improved dynamic self-adaption cuckoo search algorithm based on collaboration between subpopulations (SC-DSCS, where ‘SC’ represents ‘Subpopulation Collaboration,’ ‘DS’ represents ‘dynamic self-adaption’) is proposed. In SC-DSCS, the population of cuckoos is divided into two subgroups. The nest locations of birds belonging to the first subgroup are updated according to the traditional CS algorithm so as to retain the global search ability, and the second subgroup produces the corresponding nest locations for next generation by flying from the better nest locations to enhance the local search ability of the CS algorithm. Through collaboration between two subgroups, the problem that the local search ability of CS algorithm is not strong can be effectively solved. In order to reduce the probability of the algorithm falling into local optimum and improve the optimization precision, the SC-DSCS algorithm creates a new bird’s nest under the comprehensive assessment of the first three best bird’s nests. The new nest is added to the optimal bird’s nest sequence. In order to improve the adaptability of the SC-DSCS, adaptive step length control is adopted in the bird’s nest position updating process. Finally, nine benchmark functions are adopted to carry out the simulation experiments. The proposed algorithm is compared with particle swarm optimization algorithm, artificial colony algorithm and differential evolution algorithm. Simulation results show that the proposed SC-DSCS algorithm has better convergence speed and optimization precision.