Cruise control on a combine harvester using model-based predictive control

Combine harvesters are typically powered by a diesel engine, which drives a variable hydraulic pump. This pump delivers oil to a hydrostatic engine connected to the wheels. The hydraulic flow depends on the engine speed (of the diesel engine) and on the pump setting. Both can be controlled independently by the operator. The same travel speed can thus be reached with different combinations of engine speed and pump setting. The combine should be operated such that the diesel engine speed is as low as possible. In this way manufacturers can more easily comply with the new standards on noise and exhaust fumes. Combine harvester operators, however, often run the harvester at maximum engine speed during road transport, because this enables them to accelerate swiftly from zero to maximum speed by simply changing the pump setting. In this research, the engine speed and the pump setting are controlled actively. The objective of this control strategy is to reduce noise and exhaust fumes (by reducing the engine speed) and to simplify the operatorʹs task. Therefore, a nonlinear model of the propulsion system is derived. This model is then used in a model-based predictive controller (MPC) to develop a cruise control system. This concept is validated in practice on a New Holland combine harvester. A controller with and without engine speed minimisation is validated. The controller without engine speed minimisation represents a standard cruise control system. Its objective is to alleviate the operatorʹs task. Experiments show, however, that the engine speed can be minimised without losing acceleration performance compared to a standard cruise control system.