UCPSO: A Uniform Initialized Particle Swarm Optimization Algorithm with Cosine Inertia Weight

The particle swarm optimization algorithm (PSO) is a meta-heuristic with intelligence. It has the advantages of easy implementation, high convergence accuracy, and fast speed. However, PSO suffers from falling into local optimum or premature convergence, better performance desired. Some methods adopt improvements in parameters, initialization, topological structure to enhance global search ability PSO. These contribute solving problems above. Inspired by them, this paper proposes variant competitive called UCPSO. UCPSO combines three effective improvements: cosine inertia weight, uniform rank-based strategy. weight an form variable-period function. adopts multistage strategy balance exploration exploitation. Uniform initialization can prevent aggregation initial particles. distributes particles uniformly avoid being trapped optimum. A employed adjust individual particle’s weight. enhances swarm’s capabilities exploitation at same time. Comparative experiments are conducted validate effectiveness improvements. Experiments show that effectively improve performance.