Real-time path-constrained trajectory planning for robot manipulators with energy budget optimization

Abstract

Trajectory optimization is of great relevance for the robotics community. Many industry applications have energy limitations, such as subsea installed equipment or electric vehicles. This dissertation considers the optimization of path-constrained trajectory for robot manipulators with limited energy budget. The proposed strategy is based on a Nonlinear Receding Horizon Predictive Control (NRHPC) using a path parameterization of dimension one. The dynamic of the parameterized trajectory is governed by a predefined linear system, then an energy and a cost functions are defined and a NRHPC based on a Newton method is used to minimize the cost function in real time. A computed torque linearization scheme is considered to simplify the Newton method. The solution is also formulated in the manipulator task space permitting to be used with a kinematic control scheme, which is fundamental since many industrial manipulators only allow velocity control. The proposed solution is verified with several numerical simulations. Experimental results using a four degrees of freedom manipulator illustrate the applicability of the proposed solution.