Flame Temperature Monitoring And Control Of Rotary Kiln Based On Image Texture

Pellet rotary kiln is an important equipment of sintered pellets,the quality of clinker is produced by Pellet rotary kiln is influenced by sintering temperature.In order to improve the quality of clinker is produced by Pellet rotary kiln,generally,studying on flame temperature that it is biggest factor on affecting on the quality of clinker.We can use the methods of monitor and control to come true improving the quality of clinker.This paper establish the monitoring model that bases on characteristic parameter of experimental data and Fitting effect of temperature,this model can be effective real-time monitoring of flame temperature in pellet rotary kiln and guide the production.The maximum error of the model is 5% in the actual measurement,which can meet the production requirements.The image preprocessing is crucial to establish the optimal model of features of image texture and temperature monitoring.The flame images are collected from pellet rotary kiln are decomposed into three gray images by the method of RGB,we can get more information of flame.Then,the noise of three gray images is reduced by mean filter,this method can effectively improve the quality of images.In order to meet the isotherm requirements of flame,the target flame can be extracted by using the method of threshold segmentation for gray images by the method of RGB.In the image segmentation,taking into account the temperature of the flame has the isotherm characteristics,so this paper will use the method that according to 10 gray level as a class to divide the 80-255 gray level,so as to facilitate the extraction of the target image.At the same time,it can be more detailed study of the flame texture in the different temperatures.In order to effectively monitor the flame temperature,this paper established the monitoring model for flame temperature based on image texture.The characteristics of model are real-time performance,simple and convenient,universal.The calculation quantity of model is smaller than the method of measuring temperature by comparing color.In actual measurement,the model,the maximum error is 5%,can meet the requirements of production.In the course of graduation design,the theory of the fourth chapter has been used in the analysis of the flame texture and temperature of the candle,and has achieved very good results,which shows that the method is universal.Finally,the fuzzy adaptive PID controller is designed in this paper,which can effectively monitor the flame temperature and control the flame temperature stay at the optimum temperature to ensure the quality of the pellet.