Quasi-Time-Optimal Model Predictive Control Technology for Autonomous Pallet Picking

This paper proposes a pallet picking method based on time-optimal control addressing the high-precision control problem faced by automatic forklift in picking up pallet-loaded cargoes. In this paper, the local coordinate system is introduced, and the process of approaching the pallet and adjusting the orientation of the automatic forklift is transformed into a positional calming control problem through coordinate transformation, and the corresponding state regulator is further designed. Taking the consideration of pallet picking efficiency into account, the time-optimal performance metrics are fully investigated in constructing the optimization proposition, combined with the common quadratic optimal metrics. The control law follows the overall framework of open-loop optimization and closed-loop feedback with model predictive control, and generates control laws by rolling optimization. The control algorithm realizes the precise operation of the automatic forklift, leads which to accomplish the task of flexible pallet picking in the experiment.