1990 Volume 27 Issue 2 Pages 79-84
If a change of state in a process can be grasped visually using an image processing unit and the process can be controlled, it becomes one of the advantages of the system. This is because the image processing unit can then be used as a detecting element when a physical quantity in a process cannot be measured directly. However, generally, vagueness inevitably accompanies the data because the image is visual data. Thus, the control methods that make the best use of the advantage of the image processing were investigated.
First the change of state in a process is detected from the image data. Fuzzy control is performed on the basis of the data to control the level of granules inside the hopper. Using triangular type of the membership functions the effects of the shape of these functions, the magnitude of gains and the fuzzy reasoning methods related to response of the process are examined. Good process responses were obtained by applying the fuzzy control method.
