Whale Optimization Algorithm Based on Nonlinear Weights and Single Point Crossove

Abstract

Focusing on the problems of slow convergence speed and low search accuracy in the whale optimization algorithm, a whale optimization algorithm with nonlinear weights and single point crossover is proposed. Firstly, the algorithm introduces a non-linear weight factor in the stage of whales surrounding prey and bubble net attack to speed up the algorithm convergence; secondly, the algorithm selects several individuals randomly in the whale population for single-point crossover to increase communication between populations and to improve the algorithm to jump out the local maximum. Finally, through 12 test functions, the improved algorithm is compared with the whale optimization algorithm, particle swarm algorithm, gray wolf optimization algorithm and the whale optimization algorithm that only uses nonlinear weights and only uses single-point crossover. The experimental results show that the improved algorithm are improved significantly both in the convergence speed and the optimization accuracy.