大型炼钢炉炉温异常检测方法研究

A new Method of Large Steel Furnace Temperature Anomaly Detection

在炼钢炉内高温高压环境下,异常温度信号往往瞬间形成,在炉内气浪、异常压力鼓动等外部干扰下,温度异常信号会迅速发生非正常衰减,甚至与其它异常信号发生混淆,造成温度测量不准。传统的炉内温度异常检测方法在信号发生不可控衰减或混淆的情况下,信号分离和识别的过程存在较大困难,在外部干扰下,对温度信号的单独识别、提取效果不佳,温度检测存在较大弊端。提出一种采用约束模糊聚类算法的大型炼钢炉炉温异常信号检测方法。根据傅里叶变换方法对信号采集设备采集到的信号进行隔离处理,克服由于信号混淆带来的困难。根据约束模糊聚类算法,对隔离后的信号进行聚类,使其成为有效的识别特征,完成了异常温度的检测。实验结果表明,利用改进算法进行大型炼钢炉炉温异常检测,能够极大提高异常温度信号检测的准确率。

A fuzzy clustering algorithm is presented based on constraints of large steel furnace temperature anom- aly signal detection. According to the Fourier transform for signal acquisition device in isolation to the collected signals processing, to overcome the difficulties caused by the signal confusion, the fuzzy clustering algorithm is designed to separate the signals after clustering, make it become effective recognition characteristics, and complete the abnormal temperature detection. The experimental results show that the improved algorithm can greatly enhance the accuracy of abnormal temperature signal detection.