PARETO FRONT CHARACTERIZATION FOR MULTIOBJECTIVE OPTIMAL CONTROL PROBLEMS USING HAMILTON-JACOBI APPROACH
PARETO FRONT CHARACTERIZATION FOR MULTIOBJECTIVE OPTIMAL CONTROL PROBLEMS USING HAMILTON-JACOBI APPROACH
复制标题
DOI：
10.1137/18m1176993
复制
发表时间：
2019-01-01
影响因子：
2.2
通讯作者：
Zidani, Hasnaa
中科院分区：
文献类型：
--
作者：
Desilles, Anna;Zidani, Hasnaa
在这项工作中，我们提出了一种新的方法来刻画多目标最优控制问题的Pareto前沿。我们的方法是基于Hamilton-Jacobi-Bellman框架。我们定义了增广动力系统的辅助最优控制，证明了Pareto前沿是相应增广值函数的0水平集的子集。这些表征结果为计算整个Pareto最优前沿和相应的最优轨迹提供了一种有效的数值方法。数值算例表明了该方法的有效性。
In this work we present a new approach for the characterization of Pareto fronts for multiobjective optimal control problems. Our approach is based on the Hamilton-Jacobi-Bellman framework. We define an auxiliary optimal control for an augmented dynamical system and show that the Pareto front is a subset of the 0-level set of the corresponding augmented value function. The characterization results lead to an efficient numerical procedure for computing the entire Pareto optimal front and the corresponding optimal trajectories. Some numerical examples are considered to show the relevance of this approach.