14A-23

Y. A. KWON¹, S. J. Lee², C. G. Hong², T. S. Han², and J. Y. Kang². (1) Department of Food Science and Technology, Woosuk University, Hujung-Ri 490, Samrye-Eup, Chonbuk, 565-701, South Korea, (2) Department of Food Science and Engineering, Dongguk University, Phil-Dong, Chung-Gu, Seoul, 100-715, South Korea

For the improvement of old control method which used simple comparison of torque with main screw rpm, the algorithm of fuzzy logic controller was developed based on the general principles of extrusion. In start-up for extrusion of corn-grits as a feed material, water feed was maintained three times more than its set point value of regular control for the extrusion process. The flour feed and the main screw rpm were controlled by fuzzy reasoning to prevent a blockage phenomena. After the flour feed reached to its set point, the water feed was sequentially decreased up to its original set point. In fuzzy reasoning, input variables were torque% (represents to present torque/maximum torque) and main screw rpm% (represents to present rpm/maximum rpm). Main screw rpm was estimated by fuzzy reasoning to keep the main screw rpm% at the higher level than the torque% in excess of 20% and 30%, respectively. To manipulate the temperatures of the barrel and the die, the proportional-integral-derivative (PID) algorithm for heating and the periodic control method for cooling were used. The periodic control algorithm was that the cooler was on and off for the different time during given time, respectively, depending on the difference between the present temperature of the barrel and the set point, and the procedure was repeated periodically. Consequently, the successful automation was achieved reproducibly and economically compared to manual and old control methods without blockage.