Internal model control (IMC) of a fruit drying system

Abstract

The use of automation within the food processing industry, has allowed for the continual improvement in production and quality control, including that for dried fruit products. By monitoring and controlling process variables within a commercial fruit dryer (dehydrator), quality control features such as moisture-content, texture, and colour etc, can be attained. The three minimum process variables within most commercial fruit dryers to achieve successful drying are air-temperature, air-velocity, and relative-humidity. This project investigates the control of air-temperature in a fruit dehydrator by firstly implementing a PID controller. Then as a separate exercise, an IMC controller is implemented, and a performance comparison between the PID and IMC controllers was conducted. The PID controller was initially designed using the Ziegler-Nichols (step response method), but was determined to be too inaccurate for this purpose. So optimization using Steepest Descent Minimization was also used to determine the PID controller gains. IMC being considered a robust adaptive controller [6], is especially suited to this plant, having both a large time-constant and transport-delay. Theoretically IMC can always provide perfect system stability if the open-loop plant is first determined to be stable and an exact model of the plant is maintained, further allowing the designer to adjust other filter parameters without effecting stability.