Abstract

This paper presents a development of QFT-based self-tuning controller for a conveyor belt type grain dryer plant. Grain drying process is complex due to long time delay, presence of disturbances and plant uncertainty. QFT technique potentially has excellent solution towards this problem due to its well known capability to achieve robust performance regardless parameters variation and disturbances. The mathematical model of the grain dryer plant is obtained using system identification based on real-time input/output data.  A fixed robust controller could be designed using QFT technique; nevertheless the uncertainty range must be defined. However, in grain drying process, the parameters’ variations are unpredictable and may exceed the defined uncertainty ranges. Therefore, adaptive control with integrated Quantitative Feedback Theory (QFT) constraints is proposed to adapt larger parameters variation. Improved results are obtained by using the proposed method as compared to standard QFT procedure in terms of smaller percentage overshoot and shorter settling time when dealing with larger uncertainty range.  In addition, the design methodology of the proposed controller design (loop shaping) was improved such that the dependency on human skills was removed and the controller design was done online.

Key words: Self-tuning, quantitative feedback theory, adaptive, grain drying, system identification.