Temperature Regulation using Self-Tuned Fuzzy PID Controller for a Laboratory-Scale Fluidized Bed Ccoffee Roaster

Abstract

One of the most crucial characteristics during roasting has been the time-temperature progression of a batch of coffee beans. The roaster uses it to imitate certain flavor characteristics that occur to them. As a result, the coffee bean pile temperature trend during roasting is replicated mostly utilizing a commonly accessible PID controller. However, due to the complexity of tweaking the PID parameters, the controller is rarely maximized to its maximum performance. This article compares the performance of PID controllers tuned using M-constrained Integral Gain Optimization (AMIGO) with the proposed self-tuned Fuzzy PID controller. The controller's step responses, setpoint tracking, and disturbance rejection are evaluated. In this study, performance indicators such as percent overshoot, settling time, and integral indices such as ISE and IAE were employed. Results indicated that the proposed controller was shown to be the top-performing controller in all performance evaluations. The self-tuned Fuzzy PID controller outperformed the conventional PID controller in eliminating temperature overshoot and decreasing the settling time.