JPS58213664A - Cement manufacturing equipments - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a cement manufacturing apparatus incorporating a cement quality prediction system.

The quality of cement is represented by its 4-week (28-day) strength, but this 4-week compressive strength cannot be determined at the shipping stage.

Therefore, in reality, quality control is performed using various substitute characteristic values. The causes of variations in cement strength are diverse, but 91 quality design is extremely difficult.

The factors contributing to the variation in the 4-week strength of cement can be broadly classified into the chemical components of the raw materials, their properties and uniformity, differences in the degree of calcination, and the fineness and particle composition of the final product pulverization process.

For example, quality control at a cement factory is carried out as follows. That is, in the raw material preparation process, the chemical composition is analyzed every minute by continuous fluorescent X-ray analysis, and this data is processed by a computer to change the extraction ratio of E and N of each raw material to achieve the target component. Firing process control, which is the main cause of strength variations, is largely dependent on the operator's empirical judgment.
It is closely related to quality.

Here, for the clinker, which is a semi-finished product, the weight, unreacted amount, etc. are measured intermittently as quality check items, and the measurements are immediately checked by the operator so that they can be reflected in the process.

However, with the above-mentioned quality control system, control of the particularly important firing process depends on the operator's experience, and therefore, it is possible to ensure sufficiently stable quality control and furthermore stable cement quality with a stable design strength. It is difficult to predict and reflect this in the manufacturing process and maintain it.

This invention was made in view of the above-mentioned points.
To provide a cement manufacturing device that can predict the weekly strength and also perform adjustment operations in the suki firing process, thereby reliably maintaining cement quality with a stable design strength. This is what we are trying to do.

An embodiment of the present invention will be described with reference to the following aspects. First, this device obtains a large number of measured values in order to observe the firing situation, and these measured values are taken out in relation to the kiln system as shown in FIG. 1. In Fig. 1, 11 is a kiln.For this kiln 11, there is a 4-stage cyclone 12.3-stage cyclone J3, and a 2-stage cyclone 14.1.
Raw materials are supplied from the stage cyclone 10 and burned by the burning flame from the burner 15. In this case, the kiln 11 is rotated by a power source 16 such as a motor, and the raw material fired in the kiln 11 is led to a cooler 17 and taken out as clinker. From such a kiln system, for example, the following eight types of measurement data x1 to X8 are detected.

X 1 ''' Secondary air temperature (℃) X2...Kiln outlet clinker temperature ('C)X3...
Kiln drive motor torque
Stage ICRON raw material temperature (℃)X 7...Exhaust gas No.
x Concentration (ppm) The zero contact compensator 1
B, 19.20 respectively, data X2 and X4 are measured values obtained with the radiation thermometer as well, and
22, the data X3 is sent to the computer 2 through the analog integrator 23 and the motor torque needle signal.
It is supplied to the φ converter No. 4. Further, data X7 and X8 are directly supplied to the ψ conversion section.

Figure 2 shows the schematic configuration of a manufacturing device that performs a cement quality prediction system.
Here, data X1 to X8 are detected from the kiln system in the state shown in FIG. 1, and these data are supplied to the multiple regression equation calculation section 31. Further, this device is provided with a data acquisition unit 32 for target coefficients such as the firing status observed by microscopic observation of clinker, and the target coefficients from this data acquisition unit 32 are also supplied to the multiple regression calculation unit 31.

First, in order to obtain the target coefficient, a predetermined produced cement is sampled, and this cement sample has a mortar compressive strength Y (k
gf/m ).

Also, record the measured value when the cement is fired in the kiln. This total value is, for example, 9X as described above.
There are 8 types of 1-X 8, but if necessary, other measured values may be added or unnecessary ones may be removed to make the number of P pieces.

In this way, when at least 50 sets of data X1=-XP have been removed, a regression equation in which Y is the target variable and XI to xp are explanatory variables is calculated as follows.

(1) First, create the following data matrix and call it X.

Here, n: number of data sets p; number of measured values used xlj'', first measured value yi of the first data: 1
Target variable of th data (2) Next, let the product of the transposed matrix X' of X and X be a 95 matrix. The contents of this S matrix are as follows.

(3) Sweep this matrix S using Sp+2 and p+2 as axes to create a sum of squares and sum of products matrix S.

Here, 5XIC= (sjj') = (Σx t
/Lxl J ηj) Calculate on measured values.

Dj=SjySyj/Sjj Next, the variance ratio of the residual after incorporating Dj and its variables (measured values) into the regression equation is calculated using the following equation.

here The number of measurements already entered into the regression equation.

Therefore, the first time will be zero.

F with (5) = maximum F among (Fj) Jxmx
is selected and the matrix S is swept out with Sjm□tjmax as the axis.

(6) Repeat the calculations in (4) and (5) above, and end when the maximum value among the lower values of variables (measured values) not yet incorporated into the regression equation becomes IFIN=2.5 or less.

(7) Also, regarding the variables (measured values) that are included in the regression equation, Ten calculates the following, and if Fj<Fout=2.5, remove variable j from the regression equation. That is, S is swept out with Sjj as the axis.

(8) Finally, the matrix S becomes the following matrix.

Therefore, b=(bl,...bp) is incorporated into the regression equation (hereinafter referred to as brs, brl...
...t)rk. ζHere, k is the number of variables (measured values) incorporated into the regression equation), and the equation shown below is an estimation equation for the quality Y (28-day mortar compressive strength).

Y = b (, +brl)51 +b y2Xr2
+------+ b rkXrk--(2-1)(9)
Note that FIN used in (6) and (7) is 2.5.
, FOUT=2.5 may be respectively 1
．． It is also possible to change the value within the range of 5 to 3.5.

bo, brl, brl... obtained as above
... tlrk is stored in a predetermined storage area of the computer, and Xrl and Xrl are constantly calculated by the CPU...
Using the value of Xrk, (2) of (8)
-1) Calculate the estimated 28-day mortar compressive strength Y using the formula,
With CRT! The average values are displayed on the link every minute, every 30 minutes, and every day. Note that Xrl...Xrk refers to the current value of the kill variable (measured value) used in the regression equation.

Using the Xn-Xrk formula used in the regression equation in the current measured value obtained from the firing state detection section 30, the regression equation calculation section 3 calculates the estimated value Y of the current JIS 28-day mortar compressive strength. Aid.

This estimated value is calculated in real time by the C in the firing control chamber.
The information is displayed on the display section 33 of the RT and printer, and the kiln operation section 34 operates the kiln operation section according to the displayed content.

FIG. 3 is a flowchart illustrating the process up to the above-mentioned operation, in which a start command is given once every time from the clock to start. Then, in step 100, the current values of the measured values X1 to X from the kiln system are entered into the variable name X1-X in the resident area.

Next, in the stay and flol, it is determined whether or not there is a change in the coefficient of bO-brk and the variable type, and if there is a change, this is set in step 102. Then, Y is calculated in step f103 according to this set value. If it is determined in step 10 that there is no change, proceed to step 103.
Proceed to step 2 and calculate Y using the settings up to that point. Then, this Y is displayed on the CRT in step 104, and further stored in the time series data storage area 1 in step 105.

Next, the STETSU f106 determines whether the time is the specified "0 minutes" or "30 minutes", and if it is not, it returns to the original system as is, and when the above-determined time is reached, the STETSU yDzo7 saves the storage area. Previous 30 saved in 1
The 30 pieces of data for each minute are averaged and the average value Y30 is displayed. Then, yso is further stored in the time series data storage area 2 in step 108. At the same time, clear the data in the storage area 1 in step J09,
The data for the next 30 minutes will be imported.

Next, in step 110, it is determined whether the time is, for example, 7 am. Then, at this specified once-a-day time, in step 11, the data in the 2048 storage areas are averaged to form the daily average value Yd, and this is displayed on a CRT or the like. Then, in step 112, the data in storage area 2 is cleared. That is, 7″107 and 1
The display data of No. 11 is used for kiln control and the like.

Here, the 28-day mortar compressive strength estimation formula in the data acquisition unit 32 is periodically updated as follows.

First, the aforementioned collection of cement and the recording of the corresponding data are continued and stored in the computer as a data set.

Next, using the algorithm described above with a data set that includes new data every day, week, or month, the estimation formula is
Recalculate and change the value or variable type of b(++tlrl...kirk, and at the same time display this on the CRT and printer.

The data sets used here are as many as the new data sets, with the oldest data being discarded, or in some cases, the old data sets may also be included in the calculation.

In addition, in place of the above-mentioned JIS 28-day mortar compressive strength, the following quality judgment values determined by a microscope may be used.

This method of determining cement quality using a microscope is as follows. To prepare a sample for observation, the cement clinker is ground and passed through a 100μ sieve.

When pulverizing, avoid creating too much fine powder.

The immersion liquid is made by dropping bromnaphthalene onto methylene iodide, and the refractive index is 1.705~, which is somewhat lower than clinker mineral.
1.710. The main two mineral alitites of cement,
Observe the pellet. The observation items are the grain size of the alit,
Evaluation was made using four criteria: the birefringence of the alit obtained by using a polarizing plate, the particle size of the pellet, and the color, each in four stages.
Using the following regression equation, the quality of cement, i.e. material age 2
Determine the 8-day mortar compressive strength. For example, the regression equation is as follows.

Microscope - Judgment value = 253.46 + 6.3 s (Ant,
particle size) + 21.90 (double refraction) + 4.0
1 (particle size of pellets) + 21.51 (color of pellets) The estimated strength value obtained by the claw regression equation calculation section 31 is calculated by the setting section 3.
5, and this estimated value is multiplied by 4, and from the clinker obtained in the kiln system, a microscope judgment means 36 similar to that described above is used.
Based on the data obtained, corrective sounds may be taught as appropriate.

After passing through the silo passage time compensator 38, the above-mentioned suggested value is compared with the target value in the comparator section 39, and according to the deviation, the added amounts of corrected clinkers A and B are adjusted and the cement particle composition is controlled. Therefore, the product has a cement mortar strength of a constant ivy, and each of the above adjustment ranges is determined by the adjustment amount calculating section 40.

Based on the above adjusted actual values, the estimated value of ground cement strength is obtained by the calculation unit 41.

Incidentally, although a time delay occurs, it is possible to increase the estimation accuracy by adding this to the data after the mortar strength actual measurement value is known on the 3rd or 7th day.

That is, cement quality is represented by its 28-day compressive strength, but it takes one month to determine it. However, according to the present invention as described above, the 28-day strength of instant cement can be predicted based on the cement kill/baking conditions and the microscopic observation values of cement clinker, and by making adjustments, cement with a stable design strength can be obtained. It is possible to ensure and maintain quality.

[Brief explanation of drawings]

The attached drawings are for explaining one embodiment of the present invention.
The figure is a diagram explaining the measurement points of the kiln system, FIG. 2 is a schematic configuration diagram, and FIG. 3 is a flowchart explaining the operation up to the kiln operation. 30... Baking situation data detection unit, 31... Multiple regression calculation unit, 32... Data acquisition unit, 33... Display unit,
34...Kiln operation section.

Claims (1)

[Claims] A means for measuring and detecting various data from a kiln during cement firing, a means for calculating a multiple regression equation for this kiln firing status data, a means for estimating cement strength from the results of calculating the multiple regression equation, A cement manufacturing apparatus comprising: means for controlling the addition of plow correction clinker to the estimated value; and means for controlling the composition of cement particles according to the estimated value.