Evaluation of Fuzzy PID controller design for coffee roasting temperature regulation

The time-temperature progression is one of the most influential factors in determining the beverage's overall quality. This is because the roasted bean takes on noticeably different qualities in the cup depending on the roasting profile used, even though both varieties will have the same colour at the end of the process. This is specifically due to the fact that roasted beans take on the characteristics of the roasting profile used. Thus, the temperature profile of the coffee bean pile while it is being roasted can be replicated in a significant manner by employing a readily available PID controller. Due to the difficulty of adjusting the PID parameters, however, the controller's performance is seldom optimized. This article compares the performance of PID controllers tuned with M-constrained Integral Gain Optimization (AMIGO) to that of two Fuzzy-PID controllers with three and five membership functions. The step responses, setpoint tracking, and disturbance rejection simulation test results for the Fuzzy-PID controller indicated a significant performance improvement. This includes an average improvement of greater than 90 % in overshoot and greater than 10 % in settling time.