中文摘要：

基于现有复杂产品装配序列的特点,建立了装配体的几何可行性、零件的重新定向次数及装配体稳定性的目标函数。在原有离散粒子群算法的基础上,引入改进的进化方向算子,该算子可较为突出的改进离散粒子群算法的局部搜索能力。提出了一种混合算法,该算法在不牺牲粒子群算法的局部搜索能力和搜索速度的同时,提高其全局搜索能力,减少算法平均迭代的步数。算例表明：该混合算法具有优良的局部搜索特性及全局搜索特性,算法可快速收敛至全局最优解,可有效解决装配序列规划问题。

英文摘要：

Based on the characteristics of the assembly sequence,the geometric feasibility,the number of assembly orientation changes,and the assembly stability are chosen to be the optimization objective. Subsequently,a modified iterative method of evolutionary direction operator（ MEDO） algorithm is used to accelerate the convergence rate of discrete particle swarm optimization（ DPSO） algorithm. Then,a new hybrid algorithm MEDODPSO is proposed. The present hybrid algorithm improved the global search ability,and reduced the average iteration algorithm efficiency. The present results show that the hybrid algorithm has excellent global convergence properties and a fast convergence rate. The hybrid algorithm based on IDPSO and MEDO is efficient for solving ASP problems.