以管壳式换热器每年的总费用作为目标函数,采用含随机扰动算子的改进粒子群算法(IPOS)对其进行了优化.在优化设计模型中,采用Bell-Delaware法描述壳侧流体,优化变量选择管程数、换热管内径和外径及间距、管布置方式、封头类型、流体分配方式、密封条数、壳程压降和管程压降.对采用IPOS算法得到的优化结果与相关文献的结果进行了比较.结果表明:IPOS算法具有全局收敛、计算精度高、稳定性好的特点,并能获得约束条件下管壳式换热器的最优设计方案. Based on the annual cost of the shell-and-tube heat exchanger as an objective function, an improved Particle Swarm Optimization (IPOS) method with stochastic perturbation operator was used to optimize the model.In the optimal design model, Bell-Delaware method was used to describe the shell fluid , The number of optimization variables to choose the number of tubes, heat exchanger tube diameter and outside diameter and spacing, tube arrangement, head type, fluid distribution, the number of seals, shell-side pressure drop and tube pressure drop. The results of the optimization are compared with those of the related literatures.The results show that the IPOS algorithm has the advantages of global convergence, high precision and good stability, and can obtain the optimal design of the shell and tube heat exchanger under the constraint conditions.