MIMO variable structure controller design for a bioreactor benchmark process

In this paper, variable structure control of a bioreactor is studied. The process has two state variables named cell mass and nutrient amount, and two control inputs to maintain the state variables at their desired levels. Although the state space representation of the system seems simple, the system displays several challenges that make it necessary to develop a good flowrate (control) management strategy. Due to the plant-model mismatch, variable structure control technique is applied and it is seen that the sliding subspace is reached in finite time and the behavior thereafter is insensitive to considerable degrees of variation in the parameters and disturbances. The design is based on the nominal model and a comparison with a feedback linearizing controller is presented. The objective of the paper is to illustrate the efficacy of MIMO sliding mode control on a benchmark problem. Overall, the results with the proposed controller demonstrate the following desirable characteristics: (i) very good tracking precision (ii) small percent overshoot values and (iii) good decoupling of the process states.