JPH02184355A - Detection of crushing status of raw material in grading plant - Google Patents

Abstract

PURPOSE:To accurately detect crushing status by detecting the crushing status of a raw material for grading in a crusher from the value of both the regression coefficient in a regression formula of load between the mutual sieving machines and the regression coefficient in the regression formula of load between the sieving machine and the crusher. CONSTITUTION:The crushing status of a raw material in a grading plant for crushing and grading the raw material is detected by the combination of a plurality of crushers and sieving machines which are superposed in a cascade shape. In this case, the load of the crushers and the sieving machines is detected in a data input part 1 and a data storing part 2. Further both a regression equation of load between the crusher and the sieving machine for supplying the oversize raw material to this crusher and a regression equation of load between the sieving machine existing in the upstream side and the sieving machine existing in the downstream side are calculated from the detected value of load in a regression equation operating part 3. furthermore the crushing status of the raw material for crushing is detected from the regression coefficients of the regression equations in a crushing status detecting part 4. As a result, the crushing status of the raw material for crushing in the grading plant can be accurately detected with good reproducibility.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for detecting the crushing status of a raw material for crushing in a sieving plant that crushes and sieves the raw material for crushing.

[Conventional technology]

For example, in a raw material yard factory in the steel industry, several granulation plans I are installed for the purpose of grading raw material ore, which is the raw material for crushing. The purpose of ore sizing is to arrange the ore that will become the blast furnace insert into particle groups of appropriate size (+10 to +25 mm), thereby increasing the porosity of the charge in the blast furnace and increasing the amount of air blown. In addition, by increasing the contact area between the reducing gas and the ore, the descending speed of the charged material is increased, the amount of iron tapped is increased, and the coke consumption rate is reduced.

The smaller the particle size of iron ore, the larger the surface area, and the better its reducibility with reducible scum, but when viewed as a blast furnace charge, if there is too much fine ore, it will impede permeability and gas distribution will also be poor. It becomes uniform, controls the shelf and slip, and lowers the amount of iron tapped. Therefore, the raw material charge is both reducible and breathable ITi.
It is necessary to sort the ore to make the grain size of the ore within the optimum range, and to remove fine ore.

Recently, the lower limit of particle size of charged ore is 8 to 10 mm, and the upper limit is 25 to 30 mm. It is extremely important to strengthen the grain size of the ore and remove fine ore, which results in uniform gas distribution in the blast furnace and good contact between the charge and the gas. As a result, each time the ore is sized, heat exchange is improved and the coke ratio is lowered, and the amount of ore with a large particle size is reduced, and the time required for reducing iron ore particles can be reduced.

Conventionally, ore mined in mines is usually only coarsely crushed at the base of the mine to the extent that transportation is not inconvenient, so in granulation plan I, which has multiple crushers and sieves,
Crushing and sieving are performed.

Conventional sieving plants are, for example, disclosed in Japanese Patent Application Laid-Open No. 57-10347.
As described in , a sieving machine and a crushing machine are combined in a cassette shape. Among these, crushers were used for various types of crushing, such as coarse crushing, intermediate crushing, and fine crushing, depending on the granulation process. Note that -i indicates coarse crushing as primary crushing, intermediate crushing as secondary crushing, and fine crushing as tertiary and quaternary crushing.

On the other hand, a sieve machine sieves the crushed ore into sizes smaller than 10 meshes, and the sieve machine used in the sizing process is
There is an intermediate sieve machine that separates the supply to the crusher and the product sieve machine, and a most famous sieve machine that separates the final product into lump concentrate and fine ore.

FIG. 3 shows a sectional view of the structure of the crusher, and the operation for crushing iron ore will be explained.

Crusher (J, conical man 1 fixed to rotating shaft 30
31 and a cone cape 32 arranged around the mantle 31.

The raw material ore 33 is crushed in man 1 to 3 in the crusher room.
1 and the squeezing action of the cone cape 32. As the raw material ore 33 advances from the crusher inlet to the outlet, crushing of the raw material ore progresses.

First, the distance L (clearance) between the mantle 31 and the cone cape 32 on the exit side of the crusher is determined by the particle size of the sized product and the crushing capacity of the crusher. changes. Adjustment of this clearance value is determined by position control of the vertical movement of the mantle 31 by hydraulic movement. Therefore, regarding each crusher,
It is possible to always operate the two clearance values to the optimum value.
This becomes important in granulation plants.

Another important control object in the sizing plant is the control of the amount of ore fed, that is, the amount of cut out, based on feeder rotational speed control, which determines the throughput that goes into the sizing plant.

[Problem to be solved by the invention]

In controlling the clearance value and the cutting amount, consideration must be given to the particle size of the ore for crushing and its brand. Of these, it must be said that the particle size is constantly changing due to segregation in the stacking and poppers, which are the steps before the sizing process. When the particle size and brand of the raw ore changes, the crushing status of the raw ore changes accordingly, so the crushing status of the raw material for crushing in the crusher is accurately detected, and based on this detection result, the clearance value and the raw material It is also necessary to change the crushing power and amount of cut required for crushing the ore.

Conventionally, as a method for detecting whether or not the crusher is able to crush the raw ore to a specified particle size due to changes in the grain size and brand of the raw ore, on-site operators can detect the crushing sound of the crusher and the held controller. Sampling was carried out, such as observation of 1ζ of hair, and the crushing status of the raw material for crushing was detected based on the sampling results. Therefore, the quality of the results of detecting the state of crushing depends on the skill of the operator. As a result, the ratio of lump ore of a specified particle size to fine ore and the production amount of each are also dependent on the skill of the operator.

In order to solve such conventional unresolved problems, the present invention aims to provide a method that can accurately and reproducibly detect the crushing status of the raw material for crushing in the granulation plan I. shall be.

[Failure to solve the problem]

- In order to achieve the above object, the present invention aims to improve the crushing situation of raw materials in a sieving plant that crushes and sieves raw materials for crushing using multiple combinations of crushers and sieves stacked in a cascade F shape. In the method for detecting the state of raw material crushing in a sieving plant, the step of detecting the load of the crusher and the sieve machine, and the step of detecting the load of the crusher and the sieve machine, and applying the sieve to the crusher and the crusher from the detected load value. A step of calculating a regression equation of the load with the sieving machine to be supplied and a regression equation of the load with the sieving machine existing in the flow and the sieving machine existing downstream, and crushing the raw material for crushing from the regression coefficient of the regression equation. It is characterized by detecting the situation.

[Effect]

When the particle size of a brand of the raw material for crushing changes, the magnitude of the regression coefficient in the regression equation of the load between the sieving machines and the regression coefficient of the regression equation of the load between the sieving machine and the crusher changes. Therefore, the crushing status of the raw material for sizing in the crusher can be detected from the value of this regression coefficient.

〔Example〕

Next, one embodiment of the present invention will be described.

FIG. 2 is a flow sheet illustrating the grain grading process of an embodiment of grain grading plan 1- to which the present invention is applied.

In FIG. 2, the crushing units 1cR1 to CR4 are cumulatively configured in a cascade, and similarly, the sieving machines SCI to
The SC8 is also configured in a cascaded manner.

Raw material ore (lumps) are supplied to the granulation plant configured as described above by cutting the raw material ore from the raw material ore pile P by the reclaimer RE, temporarily retaining it in the serge pump SH, and then controlling the amount of cut out by the apron feeder AF. be done and done.

This cutout amount is measured by Hell I/Wear BWI. And steel stone block 40, which is a sized product of the raw material block.
．． Powdered ore 4] is measured by the heldware BW2BW3.

In addition, +25m each in lump steel stone 40 and powder ore 41
m ratio and -5mm ratio are weighed and evaluated by Sampler SA.

Equipment 7 such as the above-mentioned crusher and sieve machine! As shown in Figure 2, they are connected by a heel conveyor indicated by a solid line or a dotted line.

Incidentally, the Juki process consists of a complicated cascade flow including a recycling flow, and this flow can be changed in various ways depending on the operation pattern.

As the raw material ore moves from the upstream side to the downstream side of the flow explained in FIG. 2, it is crushed and sieved into lump ore having a predetermined particle size range.

In the granulation plant explained in Figure 2, the loading conditions of each crusher and sieve that make up the granulation process are as follows:
It is equipped with a current value or a detector that detects the current value, and from the detected value, a regression equation of the load of the crusher and a sieve machine that supplies the sieved raw material to the crusher and a sieve machine located upstream. The controller is provided with a controller that calculates a regression equation of the load between the sieving machine and the sieving machine located downstream, and outputs a detection result of the crushing state of the raw material for crushing from the regression coefficient of the regression formula.

A block diagram of this controller is shown in FIG.

In Figure 1, ■ is a data input section that non-pulls the current or power detection values of all the crushers and sieves of the sieving plant at regular intervals, and 2 is the data input section that is sent from the data input section 1. When the data sampled in the pig storage section 2 reaches a predetermined constant number, a data storage section 3 stores the pulling data sequentially, and when the data sampled in the pig storage section 2 reaches a predetermined constant number, the data is sent to the crusher and the crusher from this constant number of sampling data. A regression equation calculation unit that calculates a regression equation for the load with the sieving machine that supplies the sieved raw material and a regression equation for the load (current detection value or power detection value) between the sieving machine located upstream and the sieving machine located downstream. , 4 is the regression representation, and from the coefficient of the regression equation calculated in Minoh 3, CR
This is a crushing status detection unit that detects the crushing status of each of the crushers I to CR4 and outputs the contents.

The data input unit 1 samples the power or current values of all the crushers and sieves in the sieving process at a predetermined period,
It reports to the regression equation calculation unit 3 that the number has been recoded for each sampling. The sampling data is then output from the data input section 1 to the data storage section 2. Note that when sampling data, it is desirable to remove noise by -order delay processing.

The data storage unit 2 sequentially reads the sampling data sent from the data input unit 1 and stores it in a predetermined storage area. When a certain number of data are accumulated, the regression equation calculation unit 3
A certain number of accumulated data is sent from the data storage unit to the data storage unit.

After that, when new data is manually input from the data storage section 1 to the data storage section, the data storage section 2 discards the oldest data and then stores new data, always storing a certain number of the latest data. .

Each time the regression equation calculation unit 3 receives a report of execution of zumbling from the data input unit 1, it reads a certain number of data accumulated in the data storage unit 2, and then reads the data from the data storage unit 2. Data with abnormal values among the data,
In other words, data consisting of prominent values is removed and the remaining data is used as data for calculating the regression equation.

As a method for removing the above abnormal data, for example, there is a method that removes data that is more than α times the standard deviation from the average value of all the read data, and a method that sets a lower limit value for - and falls outside the range. There are methods to remove data, etc.

From the remaining data from which abnormal data has been removed, the regression equation calculation unit 3 calculates a regression equation for the loads between the sieving machines located above and in Shimoura in the grading process in Fig. 2, and between the crushing machine and the sieving machine. Calculate the regression equation for the load. Since the regression coefficient varies depending on the grain size of the raw material ore for which the regression coefficient of the regression equation is calculated, it is possible to detect or estimate the crushing state based on the magnitude of the regression coefficient or its variation.

In the data input section 1 and the data storage section 2, the sampling data is linked to a constant at a predetermined time interval. This is to detect the raw material crushing situation by determining the tendency of fluctuations in (current, power).

In the process shown in FIG. 2, the sieving process between the crusher and the sieve machine, which is not shown in FIG.
It can be divided into three patterns (1) to (3) shown in the figure.

In FIG. 4, a triangular 40 is a sieving machine, and a trapezoidal 41 is a crusher.

Sieve fraction 1sc1. shown in FIG. The combination of SC4 and crusher CRI corresponds to the pattern shown in Fig. 4 (1), i minute asc2. The combination of SC3, SC5 and crusher 1cR3 corresponds to the pattern in Figure 4 (2), and similarly the combination of sieve machines SC5, SC6, SC7 and crusher CR3 corresponds to the pattern in Figure 4 (2). Further, the combination of the sieving machines SC7 and SC8 and the crushing mcR4 corresponds to the pattern shown in FIG. 4(3).

The detected values of the load (current or power) of each crusher and sieve machine shown in FIG. be included.

The regression equation calculating section 3 calculates a load regression equation between the sieve machines and between the sieve machine and the crusher. Shown in the table.

Table 1 Regression formula calculation unit 341 calculates a regression formula for the load for each combination of each set of grain size flow patterns (1) to (3) in FIG. 4, and calculates regression formula coefficients from this same station formula. .

Next, it will be explained how this regression coefficient changes in response to changes in the original brand (hardness) and particle size.

When the ore for crushing flows between two sieves, the coefficient of the regression equation for the load of the two sieves increases as the difference in particle size distribution between the two sieves increases.

This means that 8Ji is sieved as shown in Figure 5:
C is based on the knowledge that the smaller the average particle size of the stones, the greater the fluctuation in the load (current value) of the sieving machine.

FIG. 6 shows an example in which regression analysis of load (current) was performed between sieving machines. In the graph of FIG. 6, the vertical axis is the current value of the sieve machine on the one-stream side, and the horizontal axis is the -order regression line of the current value (both units are A) of the sieve machine on the downstream side.

FIG. 6 shows the state in which the particle size distribution of the raw material ore for crushing changes from small to large.

As can be seen from the graph of FIG. 6, it can be seen that the greater the difference in the raw material particle size distribution, the greater the slope of the regression line (regression coefficient). A large slope of the regression line means that the raw material ore is well crushed by the crusher installed between the sieves. On the other hand, if the coefficient is small, it means that the crushing of the raw wood ore is poor. Therefore, by observing the magnitude of the regression coefficient of the load between the sieving machines or its fluctuation, it is possible to detect or estimate the crushing status of the raw material for sizing.

In addition, when the ore that has been sieved by a sieve machine and remains as raw material on the sieve is crushed by a crusher, the coefficient of the regression equation of the load between the sieve machine and the crusher is determined by the difficulty in crushing the raw material (the particle size is (larger, harder raw materials, etc.) FIG. 7 shows an example in which it was detected whether the crushing power for crushing raw materials changes depending on the crushability of the raw materials.

In the graph shown in Fig. 7, the horizontal axis is the current value (in A) of the second-stream sieve machine, and the vertical axis is the -th order regression line of the electric power value (unit: KW') of the downstream crusher. It is.

It can be seen from experience that the slope of the regression line (regression coefficient) becomes larger as the raw material is known to be more difficult to crush.

Therefore, by observing the magnitude of the regression coefficient between the crusher and the sieve or the variation thereof, the crushing condition] 5 can be accurately detected or estimated.

The regression formula calculating section 7 shown in FIG. 1 performs regression analysis of the load for each combination not shown in Table 1, and calculates regression coefficients.

Returning to the grain sizing process in Figure 2, (1) of 4M
The sieving flow pattern (1) shown in FIG. 2 corresponds to the combination of the sieves SCI and SC2 and the crusher CRI in FIG.

If the regression coefficient between the sieving machine SCI and the crusher CRI is large (or gradually increases), it means that raw materials that are difficult to crush are being crushed and the crusher CRI is in a crushing situation that requires crushing power. is detected.

On the other hand, if the regression coefficient between the sieve fractions asc1 and SC2 is large, the particle size distribution has changed significantly. In other words, crusher CR
I allows it to be detected that the raw material is better crushed. On the other hand, if this regression coefficient is small, the crushing is insufficient due to the crushing 1cRI, and the raw material is difficult to crush, or a defective crusher is detected or presumed.

Looking at the sieving flow pattern (2) in Fig. 4, this is due to the combination of sieving machines SC2, 3, 5 and crushing machine CR2, and the combination of sieve fractions 1sc5, 6.7 and crushing 1cR3. Corresponds to matching.

Regarding the former combination, sieve machine SC2 and SC
If the regression coefficient between 5 and 5 is large, the fracture 1icR2-(
It can be seen that the raw materials are well crushed. Also, if the coefficient of the regression equation between the sieving machine SC2 and the crushing machine CR2 is large,
It is detected that the crusher CR2 is in a crushing situation that requires crushing power.

Furthermore, the proportion of powder in the sieve material can also be detected by the magnitude of the regression coefficient between the sieve fractions 1sc2 and SC3. Therefore, from the change in the powder ratio, it is detected that the raw material for crushing is soft and is in a crushing condition that reduces the crushing power of the crusher CRI.

The combination of the latter set including the surface and the fracture 11ch'3 is also similar to the former.

Next, if we look at the I-co pattern (3),
It is a combination of sieving machines SC7, 8 and crushing machine CR4, and by analyzing the regression coefficient in combination No. 9, 104σ in Table 1, the crushing status of the raw material for sizing can be detected as described above. .

The crushing state detection section 4 shown in FIG.
Outputs the crushing status. This crushing condition consists of changes in the ore particle size distribution due to crushing and the magnitude of the crushing power required by the crusher at that time.

Furthermore, this crushing situation (particle size distribution, crushing power) can be determined from the regression coefficient.
When calculating the regression coefficient, the detected content of the regression coefficient is non-linear, and empirical values obtained through experiments in step 2 are often used to determine the crushing status such as particle size distribution, so fuzzy inference method is used. It is desirable to use it.

Currently, raw ore is being crushed in a granulating plant (not shown in Figure 2). If it were to flow into the grain process, it would be difficult to crush it in the crusher, resulting in a crushed product with a particle size that exceeds the specified particle size range, so the regression coefficient between the sieving machines would be small, as shown in Figure 6. . Moreover, as shown in FIG. 7, the regression coefficient between the sieving machine and the crushing machine becomes large.

Therefore, from the fluctuation of this regression coefficient, it is possible to detect a situation in which it is difficult to crush the raw material entering the crusher. In this case, the output content from the crushing state detection unit 4 is that the change in particle size distribution is small and the crushing power at that time is large. By controlling manually or by a microcomputer 6, 2, the clearance value of the crusher is reduced or the amount of raw material cut out is controlled to follow the fluctuations in the crushability of the raw material and compensate for the fluctuations. It is possible to continue crushing the raw ore in the same direction.

If the raw ore is well crushed by the crusher, the 6th
As shown in the figure, the value of the regression coefficient is large, and the value of the regression coefficient is maintained to be large. Furthermore, as shown in FIG. 7, the value of the regression coefficient is small and remains small.

If the raw material brand is different and a soft raw material or a raw material with a small particle size flows into the crusher, the regression coefficient shown in FIG. 7 becomes small. In this case, the output content from the crushing situation detection unit 4 is as follows:
The change in particle size distribution is small and the crushing power is small. At this time, the particle size of the product after crushing tends to be smaller than the specified particle size range, so it is possible to obtain crushed products with the specified particle size by increasing the clearance value of the crusher or by increasing the amount of raw material cut out. It can be done.

In addition, if the regression coefficient between the sieving machines and the regression coefficient between the sieving machine and the crusher fluctuates when raw materials of the same brand and particle size are flowing, it is likely that the crusher is abnormal. It becomes possible to detect.

In the above example, the nine factors for determining the regression coefficient are:
Although the current value of the sieve machine and the power value of the crusher are used, the present invention is not limited thereto, and it is also possible to use, for example, the vibration value of the sieve machine and crusher.

Furthermore, although the -order regression equation has been taken as an example, it is also possible to detect the crushing status by performing multiple regression based on each combination of a sieve machine and a crusher.

Further, although the description has been made regarding sizing and crushing of steel surfaces, the present invention is not limited thereto, and can be applied to, for example, sizing of coal.

[Effects of the Invention] As explained below, according to the method for detecting the state of rough crushing in a sizing plant according to the present invention, the load on the sieving machine and the crusher is detected, and then the load on the crusher and the crusher is detected from this detected value. Find the regression equation of the load with the sieving machine that supplies the sieved raw material to this crusher, and the regression equation of the load with the sieving machine located upstream and the sieving machine located downstream, and calculate the magnitude of the regression coefficient of this regression equation. Since the crushing status of the crushing raw material is detected from the above, the crushing status of the crushing raw material in the sieving plant can be detected accurately and with good reproducibility.

Therefore, by using the detection results obtained by the method of the present invention, it is possible to achieve stabilization and high quality of the product of the sizing process.

[Brief explanation of the drawing]

FIG. 1 is a bronch configuration diagram of a controller that implements an embodiment of the present invention, and FIG. 2 is a full sheet 1-1 explaining the grain grading process of an embodiment of a grain grading plant to which the present invention is applied. ,
Figure 3 is a cross-sectional view of the crusher configuration, Figure 4 is a combination pattern of a crusher and sieve in a sieving plant, and Figure 5 is a cross-sectional view of the crusher configuration.
The figure is a graph showing the relationship between the change in the average particle size of ore to be sieved and the variation in the load (current value) of the sieve machine, and Figure 6 shows the relationship between the change in the average particle size of the ore to be sieved and the variation in the load (current value) between the sieve machines. FIG. 7 is a graph of a -order regression line between the sieve machine load (detected current value) and the crusher load (detected electric power value). In the figure, 1 is the data input section, 2 L is the I data storage section, 3)
4 is a regression calculation unit, and 4 is a crushing state detection unit. Poetry Applicant Kawasaki Steel Stock Litigation

Claims (1)

[Claims] (1) Detecting the crushing status of raw materials in a sieving plant that crushes and sieves raw materials for crushing using multiple combinations of crushers and sieves stacked in a cascade. In the detection method, a step of detecting the load of the crusher and the sieve machine, and a regression equation of the load of the crusher and the sieve machine that supplies the sieved raw material to the crusher and the upstream from the detected load value. Raw material crushing for a sizing plant, characterized by the step of determining a regression equation of the load between an existing sieving machine and a downstream sieving machine, and detecting the crushing status of the raw material for crushing from the regression coefficient of the regression formula. Situation detection method.