Solving multiobjective Optimal Control problems in space mission design using Discrete Mechanics and reference point techniques
Solving multiobjective Optimal Control problems in space mission design using Discrete Mechanics and reference point techniques
复制标题
DOI：
10.1109/cdc.2012.6426285
复制
发表时间：
2012-12
期刊：
影响因子：
--
通讯作者：
S. Ober-Blöbaum;M. Ringkamp;Garlef zum Felde
中科院分区：
文献类型：
--
作者：
S. Ober-Blöbaum;M. Ringkamp;Garlef zum Felde
许多不同的数值方法已经发展到一方面计算最优控制问题的轨迹，另一方面近似Pareto集的多目标优化问题。然而，到目前为止，只有几种方法存在的数值处理这两个问题的组合，导致多目标最优控制问题。在这方面的贡献，我们结合联合收割机的最优控制方法离散力学和最优控制（DMOC）与参考点技术计算帕累托最优控制的解决方案。通过确定一个空间使命的多目标最优转移验证了所提出的方法。由于Pareto集的近似，结构上不同的类型的使命轨迹可以被检测到，这将不会被发现的局部单目标最优控制方法，并提供重要的信息，使命设计者。
Many different numerical methods have been developed to compute trajectories of optimal control problems on the one hand and to approximate Pareto sets of multiobjective optimization problems on the other hand. However, so far only few approaches exist for the numerical treatment of the combination of both problems leading to multiobjective optimal control problems. In this contribution we combine the optimal control method Discrete Mechanics and Optimal Control (DMOC) with reference point techniques to compute Pareto optimal control solutions. The presented approach is verified by determining a multiobjective optimal transfer for a space mission. Due to the approximation of the Pareto set, structurally different kind of mission trajectories can be detected which would not have been found by local single-objective optimal control methods and provide important information for a mission designer.