引用本文:梁海燕,任志刚,许超,言金.翼伞系统最优归航轨迹设计的敏感度分析方法[J].控制理论与应用,2015,32(8):1003~1011.[点击复制]
LIANG Hai-yan,REN Zhi-gang,XU Chao,YAN Jin.Optimal homing trajectory design for parafoil systems using sensitivity analysis approach[J].Control Theory and Technology,2015,32(8):1003~1011.[点击复制]
翼伞系统最优归航轨迹设计的敏感度分析方法
Optimal homing trajectory design for parafoil systems using sensitivity analysis approach
摘要点击 2737  全文点击 1597  投稿时间：2014-09-12  修订日期：2015-09-05
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2015.40855
  2015,32(8):1003-1011
中文关键词  翼伞  轨迹优化  时间尺度变换  参数化  优化  灵敏度分析
英文关键词  parafoil  trajectory optimization  time-scaling transformation  parameterization  optimization  sensitivity analysis
基金项目  国家高技术研究发展计划项目(“863”计划)(2012AA041701), 国家自然科学基金项目(61473253), 赛博(CYBER)协同创新中心项目资助.
作者单位E-mail
梁海燕 浙江大学 智能系统与控制研究所 工业控制技术国家重点实验室 lianghaiyan99@yeah.net 
任志刚 浙江大学 智能系统与控制研究所 工业控制技术国家重点实验室  
许超* 浙江大学 智能系统与控制研究所 工业控制技术国家重点实验室 chiaohsu@zju.edu.cn 
言金 中航工业航宇救生装备有限公司  
中文摘要
      本文对三自由度翼伞系统归航轨迹优化问题进行了研究, 采用控制变量参数化与时间尺度变换相结合的 优化算法对翼伞系统的最优控制问题进行数值求解. 该方法是基于灵敏度分析的优化算法, 将控制量以及控制量 转换时间转化为一系列参数优化问题同时进行求解. 仿真结果表明, 相对于基于两端边值优化算法而言, 灵敏度分 析法只需要正向积分进行求解, 因而具有计算简单、耗时短等优点, 其控制效果良好, 距离偏差和方向偏差均满足 实际需求, 有效地提高了翼伞系统的着陆精度, 验证了该优化算法的可行性.
英文摘要
      We investigate the control problem with three degrees-of-freedom (3DOF) for the parafoil system in the optimal homing trajectory design. The control parameterization and time-scaling techniques are employed to generate numerical solutions. The method is based on an optimization algorithm of sensitivity analysis in which control variables and their transform time are converted to an optimization problem of a series of parameters for simultaneous solving. Simulation results demonstrate the effectiveness of our computational optimal control algorithm. Instead of solving the two-point boundary value problem (TBVP) from the Pontryagin maximum principle (PMP), we propose a method that requires only the integration of the state ODE and the sensitivity system with known initial conditions for all the equations, making the computational cost much less than that in solving the TBVP. Although the dimension of the sensitivity system is higher than that of the adjoint equation of the TBVP, it still can reduce the computational cost significantly. This method can effectively improve the accuracy in landing a parafoil system and reduce the control energy consumption, showing the feasibility of the proposed optimization algorithm.