Adaptive PID control for time-varying fermentation processes

PID controller is a common control regulator for industrial processes and is widely used in chemical processes. However, conventional proportional-integral-derivative controller has been found insufficient for nonlinear and time-varying fermentation processes. In this work, an auto-tuning neuron PID (ANPID) method based on particle swarm optimization is proposed to handle the time-varying fermentation process. Specifically, the PSO method is first adopted to initialize the parameters of PID. Then, an ANPID controller is utilized to adaptively manipulate the process. The design advantage of ANPID is that its control parameters can be updated using useful historical information. Consequently, the time-varying behavior can be captured and good control performance can be achieved. Simulation results illustrate that the proposed controller can obtain better performance than traditional PID.