Abstract: After analyzing and summarizing the related research achievements of particle swarm optimization (PSO) and differential evolution (DE) algorithms, this paper proposes an improved salp swarm algorithm (iSSA) with random inertia weight and differential mutation operator for enhancing the convergence rate, computational accuracy and global optimization of SSA algorithm.Firstly, the random inertia weight of particle swarm optimization is introduced in the followers’ position update equation of SSA algorithm in order to enhance and balance the exploration and exploitation ability of SSA algorithm.Secondly, the leader position update of SSA algorithm is replaced with the mutation operator of differential evolution algorithm DE, which is used to improve the convergence rate and computational accuracy of SSA algorithm.For the purpose of checking the improvement effect on SSA algorithm by the random inertia weight and differential mutation operator, simulation experiments on some high-dimension benchmark functions are performed and comparisons between the proposed iSSA and two typical improved SSA algorithms named by enhanced salp swarm algorithm (ESSA) and crazy and adaptive salp swarm algorithm (CASSA) are performed.The experimental results and analysis show that, the proposed iSSA with random inertia weight and differential mutation operator obviously improves the convergence rate, computational accuracy and global optimization when compared with two typical improved ESSA and CASSA algorithms in the case of no any increase in the time complexity of SSA algorithm.

Key words: Differential evolution, Mutation operator, Particle swarm optimization, Random inertia weight, Salp swarm algorithm, Swarm intelligence