中文摘要：

采用粒子群算法（PSO）分析了存在初始缺陷的复合材料层合板的可靠性优化问题．在层合板结构总厚度一定的情况下，以系统可靠度最大为目标函数，对复合材料的纤维方向角和相对厚度进行优化设计．单层板失效与否应用Tsai—wu准则判断，采用一次二阶矩法（FORM）计算和评价失效概率．通过具体算例及借鉴遗传算法中的在线性能和离线性能评估准则，对粒子群算法的收敛性能进行评估，结果表明：与遗传算法相比，该方法具有操作简单、计算效率高等特点，对复合材料层合结构的可靠性优化设计十分有效．

英文摘要：

The particle swarm optimization （PSO） was applied to the reliability-based optimum design of laminated compositee was taken as the constraint condition, and thes with initial imperfections. The total weight of the laminat maximum system reliability as objective function. By using the improved PSO method, the optimum design was achieved with the ply angles and the thickness ratio as the design variables. The Tsai-Wu criterion, the first order reliability method （FORM） and the system reliability theory were utilized to evaluate the failure probability of the laminate. The online-performance and offline-performance indexes developed in the genetic algorithm （GA） were used to estimate the convergence properties of the PSO. In comparision with the GA, the PSO has a simpler procedure and higher computational efficiency. The numerical examples show that the PSO is recommendable in the reliabilitwbased optimum design of laminated plates.