Abstract: A novel trust region Particle Swarm Optimization（PSO） algorithm is proposed for solving constrained optimization problems.After particle swarm evolving each time，the trust domain search is implementing for all the particles to find better individuals，thereby the capabilities of particle swarm local searching is enhanced.The test experiments for the replenishment strategy of supplier show that the proposed solution can effectively reduce the cost of the supplier replenishment，and has good application value.

Key words: trust region, Particle Swarm Optimization（PSO） algorithm, replenishment strategy of supplier

摘要： 提出了一种改进型信赖域微粒群算法来求解带有不等式约束优化问题。粒子群每一次进化后，对所有粒子执行信赖域搜索，寻找更优个体，从而增加了微粒群算法的局部搜索能力。把算法应用于供应商补货优化，实验结果表明，该方案能够有效地减少供应商的补货成本，具有较好的应用价值。

关键词: 信赖域, 微粒群算法, 供应商补货策略