Particle swarm optimization for search precision is low and the premature convergence of the defect, through the mixing algorithm is proposed and the harmony search algorithm based on chaotic thinking hybrid particle swarm optimization algorithm. The algorithm uses the Tent map, the use of chaotic characteristics to improve the population diversity and particle traversal search, while using sound strategy for development of the solution space, the introduction of the Cauchy mutation, to help jump out of local trap particles using cloud model adaptive strategy to adjust the inertia weight. At last, the optimization algorithm is applied to reliability optimization design, simulation experiments show that the improved hybrid particle swarm optimization algorithm is better than elementary particle swarm optimization algorithm to speed up the convergence rate, and easy to fall into local minimum points.