Mechanism Surrogate Based Model Predictive Control of Hydraulic Segment Assembly Robot with Sliding Friction.

In this paper, we introduce a mechanism surrogate based model predictive controller (MS-MPC) for the motion control of a hydraulically actuated heavy-duty segment assembly robot. Heavy-duty mechanisms in construction machinery frequently employ sliding joint articulations, which exhibit complex mathematical properties that hinder the performance of real-time feedforward control with an accurate model. The mechanism surrogate used in the MPC is developed using a data-driven methodology, which can speed up the dynamics solution with 99.89% of average time savings and low error. The simulation results indicate that the proposed control method can accomplish high-precision motion control for the segment assembly robot’s lifting axis, as well as robustness across a variety of loads and friction forces. During frequent reversals, the proposed method is 80.67% more accurate than the feedforward PID controller with no load and 70.68% more accurate with a 700kg load.