ZHU Yonghong, JIANG Chao, WANG Junxiang
(School of Mechanical & Electronic Engineering, Jingdezhen Ceramic Institute, Jingdezhen 333403, Jiangxi, China)

Abstract: Ceramic shuttle kiln is a kind of commonly used batch ceramic production kiln. The detection method of its firing zone temperature is closely related to the quality of ceramic products. Hence, it is significant to optimize the temperature detection method and algorithm for ceramic shuttle kiln. This paper presents an optimized detection method and algorithm for the firing zone temperatures of a ceramic shuttle kiln based on flame image pattern recognition. Firstly, the flame images of the ceramic shuttle kiln are pretreated, and both the color feature and texture feature of the flame image are extracted by L*a*b* color space and gray level co-occurrence matrix to constitute feature vectors. Secondly, the feature vectors are input into the probabilistic neural network to be identified and classified. Finally, the comprehensive database obtained from the temperature data and the flame image is used to determine the corresponding temperature values so as to recognize the firing zone temperatures of the ceramic shuttle kiln. The experimental results show that the proposed flame pattern recognition classifier has a good flame image recognition rate. Hence, the intelligent temperature detection method proposed for the ceramic shuttle kiln is feasible and effective.
Key words: ceramic shuttle kiln; intelligent temperature detection; pattern identification; flame image; probabilistic neural network