JPH0337460B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an abnormality diagnosis method for diagnosing abnormalities in continuous casting equipment using measurement of the surface temperature of continuously cast pieces (hereinafter referred to as slabs).

If the reading on the slab surface thermometer in a continuous casting machine is significantly different from the previous trend or exhibits abnormal changes, the thermometer may be defective, the spray nozzle system may be defective, etc. Both the thermometer and the nozzle are exposed to an atmosphere of high temperature, high humidity, scale, and falling water, so malfunctions are likely to occur. If the spray nozzle is clogged, the nozzle comes off, the roll floats up, etc., it becomes impossible to control the surface temperature of the slab, and it is difficult to return to normal without input intervention. Therefore, it is necessary to detect this type of abnormality as early as possible and prevent large losses through prompt and appropriate action by operators. However, in the past, there was no suitable method for remotely monitoring such abnormalities, and workers had to directly and continuously monitor the measured values of the thermometer. Additionally, if an abnormality occurs when the worker temporarily leaves his/her seat, the quality of the slab deteriorates.

The present invention enables early detection of abnormalities in surface temperature, as well as abnormalities in continuous casting equipment and the types of abnormalities, by accumulating measurement data of the surface temperature of slabs as a time series, and processing and judging the data using statistical methods. ,
The purpose of the present invention is to provide an abnormality diagnosis method that allows an operator to take appropriate measures.

According to the present invention, in a method for diagnosing an abnormality in a continuous casting machine based on an abnormality in the slab surface temperature, the temperature of the slab surface is measured at a plurality of locations along the direction in which the slab is pulled out, and the temperature of the slab surface is measured. Temperature data is stored as a time series, and the actual measured value of the slab surface temperature and this time series temperature measurement data are compared and verified through time series analysis and statistical processing, thereby making it possible to measure the actual slab surface temperature. When the value is determined to be abnormal, the type of abnormality is determined from the previous stage temperature measurement value at the position one position before the current measurement position and the change in the drawing speed of the slab from the previous stage temperature measurement position to the current measurement position. A method for diagnosing an abnormality in a continuous casting machine is provided.

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a schematic diagram showing a portion of a continuous casting machine for carrying out the present invention. In this example,
Surface thermometer 1 for controlling slab surface temperature,
2, 3, and 4 are provided in the curved band of the slab passage. 5 is a mold, and 6 is a slab passing through a curved band.
The surface temperature of the slab 6 measured by each of the surface thermometers is input to the computer 7. Inside the computer, this temperature measurement data is accumulated as a time series and subjected to statistical processing as described later, thereby determining whether or not each temperature measurement data is abnormal. When an abnormality is determined, the temperature measurement data is compared with the temperature time series of the slab indicated by the previous thermometer, and if the trends are different, it is determined that nozzle clogging, nozzle detachment, It is assumed that an abnormality such as roll floating has occurred, and a diagnostic message is transmitted to the operator through the display section 8.

Next, statistical processing of temperature measurement data accumulated as a time series will be explained. First, the temperature data x _i
If (i=1, 2, . . . , n), the measured value generally consists of the true value x' and the measurement error ε, so it is expressed as x _i =x' _i +ε _i (1).

When temperature measurement data is time-series analyzed using an autoregressive model, x′ _i = a ₁ x _i-1 + a ₂ x _i-2 +… + a _n x _in ……(2) (where m<n) It can be expressed as Here a _i (i=1, 2,
..., m) is called the maximum likelihood estimate and is obtained by time series analysis. Therefore, the residual is ε _i = x _i −x′ _i = x _i −a ₁ x _i-1 −a ₂ x _i-2 −… −a _n x _in ……(3). This residual sequence ε _i (i=1, 2,..., n) is
If an appropriate autoregressive model is used, it can be considered white noise that follows a Gaussian distribution. The mean and variance for the residual sequence are =1/m _n 〓 ⁱ⁼¹ ε _i ……(4) σ〓=1/m _n 〓 ⁱ⁼¹ (ε _i −) ² ……(5) defined. The probability density of is given by

When the slab surface temperature is newly measured, (2), (3)
The true value x' _j and the residual error ε _j of the autoregressive surface temperature can be calculated from the equation (j=1, 2, . . . , n). Its residual ε _j
According to Tievisiev's theorem, it is determined that ε _j is normal if −3σ〓ε _j +3σ〓, and abnormal if ε _j <−3σ〓 or ε _j >+3σ〓. Note that Tievisiev's theorem here means that the density function of the random variable
(x), the average value is μ, and the variance is σ ² , it means that the inequality P(|x−μ|kσ)1/k ² holds true for any positive number k.

If it is determined that there is an abnormality, check the value on the previous thermometer and the change in the casting speed while the slab comes from the previous measurement position to the current measurement position. During this time, the system diagnoses whether the problem is with the spray nozzle or the roll, etc., and outputs a message to the operator indicating the occurrence of the abnormality and the type and degree of the abnormality. FIG. 2 is a diagram showing the procedure for implementing the present invention, which is subdivided according to steel type and casting speed, and the illustrated procedure is performed for each thermometer.

According to the present invention, by performing time-series analysis of temperature data from each thermometer and its statistical processing at the time of measuring the slab surface temperature, abnormalities in the continuous casting machine can be detected early, and appropriate measures can be taken by the operator. This can prevent large losses.

[Brief explanation of drawings]

FIG. 1 is a schematic diagram showing a part of a continuous casting machine in which the present invention is carried out, and FIG. 2 is a diagram showing the procedure for carrying out the method of the present invention. 1, 2, 3, 4...Surface thermometer, 5...Mold,
6... Slab, 7... Computer, 8... Display section.

Claims (1)

[Claims] 1. In a method for diagnosing abnormalities in a continuous casting machine from abnormalities in the slab surface temperature, the temperature of the slab surface is measured at multiple locations along the direction of withdrawal of the slab, and the temperature measurement data on the slab surface is chronologically recorded. The actual value of the slab surface temperature and this time-series temperature measurement data are compared and verified through time-series analysis and statistical processing, and the actual value of the slab surface temperature is determined to be abnormal. is characterized in that the type of abnormality is diagnosed based on the previous stage temperature measurement value at a position one position before the current measurement position and the change in the drawing speed of the slab from the previous stage temperature measurement position to the current measurement position. A method for diagnosing abnormalities in continuous casting machines.