高炉风机故障预报系统

A Faults-Forecasting System for Fans of a Blast Furnace

根据长治钢铁 (集团 )有限公司炼铁厂生产的需要 ,针对其关键生产设备六高炉风机建立了故障预报系统 ,详细阐述了预测技术的选择和预测模型的建立过程 ,成功地将灰色加一元线性回归模型应用于风机运行状态的预测 。

In order to forecast the running state of fans ,a fault forecasting system was established . Through comparison of several common methods (gray system method, time series analysis method, etc.), a new forecasting model of gray system method blending with the method of linear regression with one parameter was brought forward. During modeling, the model was brought down to two parts: determinate and stochastic. To the determinate part, gray system method and the method of linear regression with one parameter were analyzed and four kinds of forecasting functions were established:Y=aX+b;Y=aX c+b;Y=a lg X+b;Y=a e cX +b. To the stochastic part, gray system method was applied to forecast the differential value of forecasting data and known data of the determinate part. The last forecasting data were the sum of the determinate forecasting part and the stochastic forecasting part. According to the variance of differential value of forecasting differential value and the known differential value above, the predictable region is given by probability 1. In addition, the selection of optimal models was discussed. Finally, the idea of programming was presented . Applications show that, this method is feasible .