Application of Model Predictive Control in Discrete Displacement Cylinders to Drive a Knuckle Boom Crane

In this paper, two Discrete Displacement Cylinders (DDCs) are used to drive the boom of a knuckle boom crane. DDCs operate by connecting one of several available pressure levels to each chamber in order to produce different forces. A trade-off exists with such systems, between the accuracy of tracking and energy dissipation due to switching. A popular way to approach this problem is a Force Shifting Algorithm (FSA). However, in this paper, the trade-off is managed by use of a Model Predictive Control (MPC) algorithm. The tracking accuracy and energy efficiency of the MPC and FSA strategies for DDCs are compared to a PID strategy for standard cylinders. The comparison is obtained by use of a computer simulation of a knuckle boom crane performing a realistic load cycle. The load cycle consists of the crane extending to pick up a load and then retracting to place it at an appropriate location. The main results show that MPC can deliver smoother and more accurate motion than FSA, while using less energy. Compared with standard cylinders and PID control, MPC uses less energy, but due to the switching of chamber pressures, the motion is smoother with the standard strategy. Both FSA and MPC can have degraded performance when a large change in load is introduced.