Recognition method of cement rotary kiln burning state based on Otsu-Kmeans flame image segmentation and SVM

Abstract

Abstract The accurate recognition of the burning state of cement rotary kiln is of great significance to improve the qualified rate of clinker quality, ensure the safe operation of rotary kiln and reduce the production coal consumption of enterprises. In this paper, a recognition method of the burning state of cement rotary kiln based on Otsu-Kmeans flame image segmentation and support vector machine (SVM) was proposed. Firstly, the flame image of the clinker burning core region in the rotary kiln was obtained by using advanced vision detection technology, and then the Otsu-Kmeans image segmentation method was used to realize the effective segmentation of the target region of a flame image. The target region of flame image includes radiation region, flame burning region and high-temperature region. On this basis, 10 feature parameters of the target region were extracted as state recognition features. Then, with the one-minute statistical average value of the extracted feature parameters as input, and the three typical burning states of the rotary kiln as output, an SVM-based recognition model for the burning state of the rotary kiln was established. Finally, the established SVM state recognition model was used to classify the flame image, and the two parameters in the model were optimized by the K-fold cross-validation K − C V algorithm. The results show that for 30 sets of test samples, the SVM-based burning state recognition model can correctly identify 28 sets of samples, which provided a feasible method for the recognition of the burning state of cement rotary kilns.