PROBLEM TO BE SOLVED: To properly predict abnormal furnace conditions by detecting blasting pressure of pressure in the furnace, frequency analyzing time series pattern and monitoring the neighborhood of a specific frequency. ;SOLUTION: In a blast furnace 10, a sensor 15 for detecting the blasting pressure and a sensor 16 for detecting the pressure in a shaft part are fitted, respectively and data detected with these sensors 15, 16 are periodically collected and stored in a data collecting part 25. In a time series frequency analyzing part 26, the time series frequency analysis of the data in the past fixed period is executed from the newest values of the collected and stored pressure data. In the case of executing this analysis with a Fourier transformation, the change with time of the frequency components in the sensor data can be obts. by executing the Fourier transformation related to each time in the time series data. The analyzed result is displayed on an observing monitor 27. At this time, the tendency to the aggration of the furnace condition is predicted in advance by observing to power near the prescribed frequency (e.g. 2.8×10-3Hz; the frequency component accompanying the charge of raw material).;COPYRIGHT: (C)1998,JPO
展开▼
