Differential evolution optimization applied in multivariate nonlinear model-based predictive control

Model-based Predictive Controllers belong to a class of digital controllers which are used in many industrial applications. Such controllers have the main advantages of dealing with optimization subject to constraints and multiple-input, multiple-output systems. The optimization of the control system behavior is based on an explicit mathematical prediction model of the plant. Usually, linear approaches are used for the prediction model. However, for nonlinear plants, linear models may limit the control loop performance or even cause instability. In this work, a Nonlinear Model Predictive Controller, with optimization based on a nonlinear model performed with a Differential Evolution algorithm, was tested for position control of a robotic arm model. Simulation results show that the utilized control algorithm was able to deal with multivariable nonlinear optimization in the presence of process constraints.