Function Optimization of Two Improved Particle Swarm Algorithms

Abstract

Aiming at the problem that in the later stage of the traditional Particle Swarm Optimization (PSO) algorithm, the particle velocity is dominated by the cognitive and social components, leading to a decrease in velocity and inertial effect, and making it difficult for particles to escape from the local optimal solution, this paper proposes two improved particle swarm optimization algorithms, namely the Adaptive Weight Particle Swarm Optimization (IPSO) and the Particle Swarm Optimization Combined with Chicken Swarm Optimization (PSOCSO). The specific strategies are as follows: First, a specific crossover operation is introduced to improve the global search efficiency of the solution, which effectively enhances the global search ability and convergence speed of the algorithm. Second, the PSO algorithm is combined with the Chicken swarm Optimization (CSO) to avoid the situation, where particles gather at the local optimal value prematurely and thus lose the ability to explore the entire search space. Through experimental simulations, a comparative analysis of the algorithm performance is conducted from three aspects: solution accuracy, convergence speed, and problem dimension. The results show that the optimized Adaptive Weight Particle Swarm Optimization (IPSO) algorithm exhibits significant advantages in performance compared with the other three contrastive particle swarm optimization algorithms and has stronger robustness.