提出一种改进的多目标粒子群优化算法，应用于飞机环控/发动机系统的综合优化. 将不同飞行阶段系统总熵产最小视为不同的目标函数，建立了多目标优化模型. 进而在基本多目标粒子群优化算法基础上，引入跳转操作、族群概念和一种全局最优位置分配方法，提出了一种改进算法，测试结果表明该算法性能良好. 采用该算法对多目标优化模型进行计算，得到收敛且分布均匀的非劣最优解集，为飞机系统综合优化提供一种新思路.

An improved multi-objective particle swarm optimization algorithm is proposed and applied to the integrated optimization of environmental control system and engine of aircraft. A multi-objective optimization model is established by viewing the total entropy generation minimum at different flight phases as different objectives. Based on the basic multi-objective particle swarm optimization algorithm, an improved algorithm is proposed through a jump and cluster operation, and a new global optimal position distribution method. The simulation results show the good performance of the algorithm. The proposed algorithm is used to compute the multi-objective optimization model, the obtained pareto optimal set has good convergence and diversity. It provide a new way for integrated optimization of aircraft systems.