Design of centralized robust pi controller for a multivariable process

A simple robust control technique based on centralized Proportional Integral is proposed for a multivariable process. The study is undertaken to improve the performance of the multivariable system in presence of plant model mismatch. The method considers Steady State Gain Matrix and process parameters like time delay and time constant for the controller design. The robustness of the proposed technique is analysed with perturbation in all the three parameters namely process gain, time constant, and time delay. The servo and load disturbance rejection response is obtained. Also, the multiplicative uncertainty is checked with spectral radius criterion. The Integral Absolute Error values of main effect, interaction effect, and uncertainty in process parameters are also compared with the centralized controller designed based on Steady State Gain Matrix. The tuning parameters for the original plant and perturbed plant is kept constant throughout the analysis. Finally, the proposed control algorithm is validated using pilot plant binary distillation column.