Nonlinear model predictive control of a heavy-duty hydraulic bulldozer blade

Abstract

In this paper, we present a study on nonlinear model predictive control (NMPC) of a bulldozer blade actuated by hydraulic cylinders. We design the controller in the manipulator joint space to track the desired joint angle trajectory. We specify the desired blade cut edge trajectory in the world reference frame and map it to the joint space. In the mapping, we consider the vehicle body orientation and generate a preview of the desired trajectory for the NMPC by estimating the future orientation based on its current value. We test the performance of our controller in MATLAB Simscape. The results show that the proposed controller tracks the desired world frame position with sub-centimeter accuracy regardless of the vehicle body motion and inclination.