基于粒子群算法的可重复使用液体火箭发动机减损优化问题研究

Particle Swarm Optimization Algorithm Based Damage-mitigating Control for Reusable Liquid Propellant Rocket Engine

针对可重复使用液体火箭发动机起动过程,研究了系统控制输入序列的变化对以涡轮叶片为代表的关键部件累积损伤的影响。在系统性能变化不大前提下,为使瞬间动态过程中涡轮叶片损伤累积较小,从而提高可重复使用发动机工作寿命和使用次数,将系统性能函数和涡轮叶片累积损伤量设为目标函数,应用粒子群算法对系统控制输入序列进行优化设计。仿真结果表明:1)在系统性能函数值相同情况下,控制输入序列引起的涡轮叶片累积损伤大不相同;2)选择优化后的控制输入序列能够实现发动机高性能条件下,降低涡轮叶片累积损伤的目的。

This paper studies the problem of tradeoff optimization between damage of reusable liquid propellant rocket engine＇s turbine blade and control input sequences during the start-up proeess. The objeetive is to reduee the damage of turbine blades without insignifieant loss in the dynamie performance, and extend the life of the rocket engine. So in this paper, system performanee function is selected as the objective function, psrtiele swarm optimization algorithm （PSO） is used to optimize the eontrol input sequence. The results of simulation computation show that： 1）different control input sequences, which desire the equal system performance function value, can generate absolute different damage; 2） the seleeted eontrol input sequences from optimized ones in start-up proeess ean relatively reduee the damage accumulation without insignificant loss in the dynamic performance.