关键词: 无约束优化, 非单调线搜索, 微粒群算法, 信赖域, 全局收敛性

Abstract:

A class of non-monotone trust region algorithm for unconstrained optimization problems is proposed.And a trust region algorithm based on particle swarm is proposed which combines the good search performance of PSO and the good overall convergence of trust region algorithm.Under appropriate conditions,we prove the global convergence of the algorithm.

Key words: unconstrained optimization;non-monotone;PSO;trust region;global convergence