The New Jersey Institute of Technology's
Electronic Theses & Dissertations Project

Title:	A successive approximation algorithm to solve motion planning problems
Author:	Chang, Cheng-Ming
View Online:	njit-etd1993-065 (x, 83 pages ~ 2.4 MB pdf)
Department:	Department of Electrical and Computer Engineering
Degree:	Master of Science
Program:	Electrical Engineering
Document Type:	Thesis
Advisory Committee:	Zhou, MengChu (Committee chair) Kuo, Marshall Chuan Yung (Committee member) Shi, Yun Q. (Committee member)
Date:	1993-01
Keywords:	Robots -- Motion -- Planning Robots -- Motion -- Automatic control
Availability:	Unrestricted
Abstract:	A successive approximation algorithm is presented to solve motion planning problems and optimal control problems. A motion planning problem is included in an optimal control problem and due to the heavily nonlinear and coulped differential equations describing the dynamics of the manipulator and several constraints imposed on the system, few efficient methods exist to solve this kind of problems. The successive approximation algorithm can decompose the global nonlinear problem into several nonlinear subproblems and each subproblem is solved by the nonlinear programming method with a highly convergent rate. Because the limited constraints and variables are included in each subproblem and only the trajectory and control sequence are needed to be stored in each iteration, the algorithm can be very easily implemented on the motion planning problems. The convergence and optimality of the algorithm are also discussed and the implementations on motion planning problems and optimal control problems are presented with excellent results. The computational results show its promise for future applications.