*** ABSTRACT 9660214 Kumar This Small Business Innovation Research Phase I project evaluates the feasibility of creating a generic, user-friendly optimal robotic trajectory planning software that is robust, fully automated, and capable of multi-degree-of-freedom optimization with path constraints. Currently, no commercial software exists which can solve generic robot optimization problems in a turn-key fashion. A differential inclusions technique recently developed at Analytical Mechanics Associates (AMA) is ideally suited to solve the robot trajectory optimization problem because the controls appear linearly. Proving concept feasibility implies: (a) construction of a generalized symbolic algorithm that creates the equations of motion of multi-degree-of-freedom manipulators from user-provided mass and geometry data; (b) representation of above equations in the differential inclusion format; (c) evaluation of analytical partial derivatives of the resulting hodograph, boundary, and path constraints with respect to the generalized states; (d) evaluation of convergence robustness of the differential inclusion approach for the entire class of open-chain manipulators and comparison to classical variational and collocation techniques; and (e) obtaining optimal solutions for high-degree-of-freedom manipulators via a special dense-sparse discretization approach. Commercial applications include performance enhancement of manufacturing processes such as spot welding, painting, and assembly, where minimizing time and energy requirements map into millions of dollars saved in direct manufacturing costs. ***
