Model Predictive Control of Self-Propelled Work Machines with Spatial Disturbance Preview

Self-propelled work machines reach greatest efficiency in terms of cost and energy at high engine loads. The required engine power to perform a task with a given velocity may vary rapidly with changing local conditions. Using forward-looking sensors, one is able to detect such changes in advance. This paper presents a method for making spatial sensor data usable for the temporal Model Predictive Control (MPC) problem. In addition, practical issues like offset-free tracking and constraint handling are discussed. The advantage of disturbance preview MPC compared to common feedback MPC is shown in simulation.