Optimization problems in multivariable fuzzy predictive control (Q707234)

The basic concepts of model predictive control (MPC) are recalled, and the control strategy based on branch-and-bound (B\&B) method is outlined. To alleviate the computational burden of the B\&B algorithm applied to the MIMO (multi-input/multi-output) systems, and to diminish chattering effect the fuzzy predictive filter is added into the control scheme. Proposed filter includes an ''adaptive'' set of incremental control alternatives (which represent the space of available control actions at current time) and a fuzzy predictive decision-maker to scale these controls. The scaled/filtered controls are then passed to the B\&B algorithm. The proposed MIMO fuzzy MPC scheme is applied to a realistic model of a container gantry crane. Numerical simulations show that the proposed method always decreases the error and takes less computational power than the strategy based on the multivariable B\&B MPC scheme.