Control of Multi-scale Dynamics System

This paper addresses the importance of the multi-scale modeling and its growing applications in many industrial processes. An idealized multi-scale model that represents a typical product inhibition process that is commonly encountered in fermentation system is developed. Based on this idealized model, a simple multi-scale control strategy is proposed and its performance is compared with that of single macro-scale control approach. Simulation study shows that for single macro-scale approach, the closed-loop performance and robustness is strongly related to the micro-scale dynamic. More importantly, the proposed multi-scale control strategy shows superior performance than that of macro-scale control approach in the face of the worst macro-scale disturbance occurrence. Additionally, this paper shows that the use of complex control algorithm such as nonlinear controller can be avoided if the right control structure can be identified and implemented.