We study the convergence properties of an overlapping Schwarz decomposition~algorithm for solving nonlinear optimal control problems (OCPs). The approach decomposes the time domain into a set of overlapping subdomains, and solves subproblems defined over such subdomains in parallel. Convergence is attained by updating primal-dual information at the boundaries of the overlapping regions. We show that the algorithm exhibits local linear convergence and that the convergence rate improves exponentially with the overlap size. Our convergence results rely on a sensitivity result for OCPs that we call "exponential decay of sensitivity" (EDS). Intuitively, EDS states that the impact of parametric perturbations at the boundaries of the domain (initial and final time) decays exponentially as one moves into the domain. We show that EDS holds for nonlinear OCPs under a uniform second-order sufficient condition, a controllability condition, and a uniform boundedness condition. We conduct numerical experiments using a quadrotor motion planning problem and a PDE control problem; and show that the approach is significantly more efficient than ADMM and as efficient as the centralized solver Ipopt.
翻译:我们研究了一个重叠的Schwarz分解分解 ~algorithm 的趋同特性,以解决非线性最佳控制问题(OCPs) 。 这种方法将时间域分解成一组相重叠的子域, 并同时解决为这些次域定义的子问题。 通过在重叠区域的边界更新原始- 双向信息, 实现了趋同。 我们显示, 算法显示本地线性趋同, 趋同率随重叠的大小而成倍提高。 我们的趋同结果取决于我们称之为“ 敏感度的加速衰变” (EDS) 的 OCPs 敏感结果。 直觉地说, EDS 指出, 域边界( 初始和最终时间) 的参数过错影响随着向域的移动而急剧衰减。 我们显示, EDS 持有非线性 OCPs, 处于统一的第二阶线性充分条件、 可控性条件和统一约束性条件之下。 我们使用夸德罗托尔运动规划问题和 IDE控制问题进行数字实验。 从直觉看, 显示, 集中式的方法比 ADM 有效, 解为 。