PID controller design for industrial beer filtration

This paper investigates the optimization and control of a beer filtration process via closed-loop PID control. The authors' previous work has found that the beer filtration process changes in both dynamics and steady-state gain, resulting in different linear models obtained at different operational times. This paper addresses the issue of multiple linear models for a plant by taking the average of their frequency responses to obtain a single set of frequency response data. It also proposes to use a PID controller design technique based on two frequency response data points. The PID controller is then validated using Nyquist loci for measuring gain and phase margins. Closed-loop control simulations that mimic the filtration process have been used to demonstrate the feasibility of maintaining the desired operating conditions of the filtration process by manipulating the system input.