烧结机尾特征断面图像采集算法的研究及应用

Research and Application of Collection Algorithm for Characteristic Section Image of Sintering Machine Plant-Tail

对机尾特征断面的出现过程进行了阐述和理论分析,提出了机尾特征断面出现的2种方式和台车临界值概念,并给出了台车临界值的计算方法。针对东烧360 m2烧结机的实际情况,克服了现有特征断面图像采集算法的缺点,研究了一种新的简单、高效、适应性强、易实现的算法。新算法通过合理地设定判断区域的上下边界和3个简单的判断阈值,实现了在各种烧结工况下都能够准确地采集到特征断面图像,并且通过对最低平均温度初值的设定,有效地解决了对辉光图像的误判问题。系统自2007年1月运行以来,特征断面图像的准确采集率达到95%以上。

The definition of characteristic section image of sintering machine plant-tail is thoroughly described and deeply analyzed for the first time, two types of characteristic section image and pallet critical value are proposed, and the calculation method of pallet critical value is demonstrated. For 360 m2 sintering machine at Donganshan sintering plant, the disadvantages of existing collection algorithm were eliminated by using, a new algorithm,which is simple, efficient, adaptable and easy to use. By resonable presetting up and down borderlines of the judging area and three simple judging thresholds, the new algorithm can exactly collect characteristic section image for every operation mode, and by presetting minimum average temperature the new algorithm can efficiently eliminate the misjudgment for glow image. Since the system was put into production in Jan, 2007, the rate of exact collecting characteristic section image has been over 95%.