Simulate anneal-particle swarm optimization（SA-PSO） is a simple and efficient stochastic global optimization algorithm

and simulated annealing is involved into particle swarm optimization（PSO） with Crossover and Gaussian mutation.The reliability optimization model maximize reducer driven system is established by combining the SA-PSO with the theory of reliability optimization design

and the reliability optimization method for cylindrical gear reducer is presented on the basis of the SA-PSO.This optimization method is used to optimize a cylindrical gear reducer

and the optimal results are much better than those by the PSO

the mixed penalty function method and the traditional design method.The example shows that this reliability optimization design method is effective

and has the predominant capability of global optimization and convergence precision.Moreover

it has a certain reference significance for other mechanical parts reliability optimization.