摘要
文章以数值仿真为手段,研究了气囊缓冲登陆系统的冲击动力学多目标优化问题。首先,以类似于"猎兔犬"着陆缓冲系统为对象,确立了以囊内初始气压与收缩绳刚度为设计变量,以缓冲系统的首次冲击最大过载和囊体织物最大应力为目标函数的多目标优化问题。随后基于D最优试验设计,采用移动最小二乘(Moving Least Square,MLS)构建了各个目标函数的代理模型,并对目标函数随设计变量的变化规律进行了探讨。最后,采用遗传算法完成了缓冲系统的冲击性能优化,并给出设计空间中关于最大过载与织物最大应力的Pareto前沿。研究结果表明,MLS模型优化算法十分适用于解决非线性程度较高的冲击动力学优化问题,并且在替代真实模型仿真计算时不仅具有较高的近似精度而且具有高速的分析效率。
Multi-objective optimization of airbag cushion landing system for deep space exploration is investigated in this paper. Firstly, the cushion landing system of ＇beagle＇ is taken as an example,the initial pressure in airbag and the stiffness of shrinking ropes are chosen as design variables, and the multi-objective optimization of airbag cushion landing system is established by chosen the maximum overloading of the first landing and the maximum stress of the fabric as the objective functions. Then, meta-models method is introduced here to construct the relationship between design variables and overloading of the landing system by use of D-optimal experiment design method, and the variation of the objective function with design variables are studied. Finally, the result of optimization for airbag cushion landing system is presented,giving the Pareto forefront. Numerical results show that meta-models method is suitable to solve the problem of highly nonlinear features such as the optimization of cushion landing system＇s response, and has not only high approximation accuracy but also high efficiency of fast analysis .
出处
《航天返回与遥感》
2012年第5期1-8,共8页
Spacecraft Recovery & Remote Sensing
基金
国家"863"计划(2008AA12A205)
南京航空航天大学基本科研业务费专项科研项目(NS2012015)
关键词
气囊
缓冲
多目标优化
冲击
移动最小二乘
D最优试验设计
深空探测
airbag
buffer
multi-objective optimization
impact
moving least square
D-optimal experiment design
deep space exploration
