Improving Coffee Roast Yield Consistency with Self-Tuned Fuzzy-PID Controller

Abstract

Precise temperature control is essential in industrial processes like coffee roasting, where consistent quality is crucial. While traditional Proportional-Integral-Derivative (PID) controllers are widely used for temperature regulation due to their simplicity, they often struggle with dynamic process changes, leading to variability in product quality. This research investigates the Self-Tuning Fuzzy PID (STFPID) controller, which uses fuzzy logic to dynamically adjust its parameters, improving performance. The study aims to compare the effectiveness of conventional PID controllers and STFPID controllers in maintaining consistent roasting conditions and product quality. Experiments were conducted with both controllers to regulate the temperature during coffee roasting, focusing on indicators such as weight loss percentage, Agtron values (indicating roast level), and peak wavelength (reflecting color properties). Time-temperature profiles for multiple batches were analyzed to assess consistency and stability. Results show that the STFPID controller significantly outperforms the PID controller, achieving a 76.5% improvement in weight loss consistency, a 71.2% reduction in Agtron value variability, and a 64.4% enhancement in peak wavelength stability. These findings demonstrate the STFPID controller's superior ability to adapt to process variations, maintaining uniform conditions and enhancing overall product quality in coffee roasting applications. The significance of this study lies in its potential to enhance the consistency and quality of coffee roasting, providing valuable insights for industrial applications and contributing to the advancement of process control technologies in this field.