Abstract
In recent years, our life is highly based on industrial applications. The majority of industrial systems processes are Multi Input, Multi Output (MIMO) systems which are a very challenging domain of research. The most crucial problem in MIMO industrial systems is how to make it robust in face of uncertainties. Besides, we find these days a huge growth in the use of non-integer order systems given their efficiencies towards uncertainties and complex systems. In control of multivariable plants there are many techniques used to react with these systems as QFT (Quantitative feedback Theory) design. The QFT method is used to convert the MIMO process into MISO (Multi Input, Single Output) sub-processes and after that we must determine the filters and controllers that guarantee the robust control against uncertainties. In this paper, a design of an automated strategy based on auto-tuning of non-integer-order pre-filter and fractional-order controller is presented for MIMO processes unlike the classic QFT approaches based on manual design to extract adequate filter and controller. A detailed design and optimization of the fractional-order PID controller based on bi-objective evolutionary algorithms is designed to control multivariable plants followed by a nonlinear optimization of fractional-order pre-filter used to guarantee our system path tracking and the bi-objective desired performance specifications. Finally, our proposed methodology is tested to prove its efficiency through application onto an industrial MIMO process.