Optimal Feedback Control of a Bioreactor with a Remote Sensor

Sensors used to monitor bioreactor conditions directly often perform poorly in the face of adverse nonphysiological conditions. One way to circumvent this is to use a remote sensor block. However such a configuration usually causes a significant time lag between measurements and the actual state values. In this work the problem of implementing feedback control strategies for such systems described by nonlinear equations is addressed. The problem is posed as an optimal control problem with a linear quadratic performance index. The linear control law so obtained is used to implement feedback. A global linearization technique as well as an expansion using Taylor series is used to effectively linearize the nonlinear system and the feedback is subsequently implemented.