Trajectory optimization of COmpliant HuMANoid (COMAN) robot arm using path parameter based dynamic programming

In this paper, the problem of trajectory optimization for the COmpliant huMANoid robot (COMAN) arm is solved by Dynamic Programming. It is based on dimensionality reduction and uses a path parameter to convert the dynamical equations of the robotic manipulator to a set of second order differential equations. First, the path parameter is calculated from the given geometrical path information and then the path parameter based reduced dynamical equations and the terminal conditions are described in the discretized form. Bellman's optimality principle is applied to calculate the minimum performance index and to optimize the objective function. To show the effectiveness of path parameter based Dynamic Programming approach, it is tested on a simulation of the COMAN robot arm and the obtained optimal results are compared in terms of numerical accuracy and computational time with the optimal solutions obtained by other state-of-the-art techniques.