Abnormal State Diagnosis of Sintering Image Based on SVM

Abnormal sintering state is often caused by the changes of improper operation in the sintering process of rotary kiln. If not addressed immediately, control system performance will be deteriorating, and even the crash will be caused. Current approaches of pattern recognition cannot be applied immediately to recognizing such abnormal sintering state of rotary kiln. Therefore, integrating both image processing method and support vector machines(SVM), this paper studies a new and enhanced approach on state recognition of abnormal sintering image, namely,image pretreatment, image segmentation, features extraction, automatic choice of SVM parameters and abnormal state diagnosis technology of sintering image in rotary kiln. Finally, the experimental results show the effectiveness of the approach.