In targeting the production process of the vulcanizing accelerator, a newly developed granulation control system is proposed for saving labor in the process and improving the quality of the product. The system is assembled referring to the expert's procedure for the regulation of the granulating state that is made up of two images about the shape of pellets and the state of the granulating machine. To cope with these tasks, the proposed system includes the following two independent feedback controls. The shape of pellets is regulated through an image processing and a combined estimation with a neural network and a fuzzy model. Similarly, the state of granulating machine is also regulated by using fuzzy inference. Through the examination of the experimental results concerning the accuracy of the models and the following of the assembled system, it is confirmed that the system proposed is useful for practical applications.
