From boundary lines to boundary surfaces for dynamic programming with final state constraints

W. Cottin, Y. Liu, G. Colin, A. Charlet, S. Houille

IFAC PAPERSONLINE(2023)

引用 0|浏览0
暂无评分
摘要
This article addresses the generalization of the boundary lines method for Dynamic Programming, called here boundary surface. This method can be applied on any system with 2 states as long as there are border constraints (initial or final). The method has been applied to two different problems. The first one is a simplified convex acceleration problem with final constraints on speed and position. The second problem is the energy minimization of a parallel hybrid electric vehicle along a trip in a non linear and non convex formulation. Results show a non negligible accuracy improvement despite the increase in calculation time. This can yield more accurate optimal results than those found by any possible mesh grids on an average computer for a reasonable calculation time. Copyright (c) 2023 The Authors.
更多
查看译文
关键词
Dynamic Programming (DP),Boundary Lines (BL),Boundary Surface (BdS),Optimal Control Problem ( OCP),Energy Management Strategy (EMS)
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要