JPH0924291A - Control method for powder producing method utilizing partial classification ratio model and powder producing apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To compose a system by which the optimum operation of a process to pulverize powders and classify the resultant powder into a plurality of levels is carried out by producing production yield and a partial classification ratio model based on the particle size distribution before classification and the measured value of the particle size distribution of products, estimating the production yield and the particle size distribution of products by utilizing the partial classification ratio model, and control the variables of the process as to carry out the process in the optimum conditions. SOLUTION: This control method is to control the production of powders involving pulverization and classification processes: a partial classification ratio model is computed based on particle size distribution before and after classification, respectively, and optionally set parameters of the partial classification ratio model are changed. The parameters of the partial classification ratio model with which a desiring particle size distribution and yield ratio after classification can be achieved are determined and process variables are computed based on the determined parameters of the partial classification ratio model while utilizing database in which the parameters of the partial classification ratio model and the process variables are correlated mutually.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of controlling a system for pulverizing powder and classifying it into a plurality of pieces, and an apparatus therefor.

[0002]

2. Description of the Related Art Conventionally, in a method (process) for producing powder such as cement, fertilizer and wheat flour by granulating, crushing and classifying, the particle size distribution has been used as an index for operation control. Then, process variables of the crushing process and the classification process so as to obtain a desired particle size distribution (for example, the number of revolutions of the rotor of the impact crusher and the conditions such as the gas flow rate of the multi-product simultaneous classifier at the time of crushing or classification). Control by operating. This is because of changes in the quality of raw materials,
This is because the particle size distribution of the product changes due to changes in the powder flow rate in the classification process, temperature, humidity, and changes over time in the process. In the conventional automatic control or manual control, a desired particle size distribution is compared with an actually obtained particle size distribution, as described in, for example, Japanese Patent Application Laid-Open No. 6-95705 "Control device for granulation system". The process variable was directly manipulated from the numerical value of. Most of these methods manage operation based on experience and experience.

[0003]

However, there is a growing demand for the particle size distribution of products such as toner for developing electrostatic images in recent years, and there is a problem in operating process variables based on conventional experience. It started to occur.
For example, when making the particle size distribution a desired particle size distribution by manipulating process variables, there are multiple manipulating process variables that give the same effect, and it was not clear which process variable was optimal to achieve. . That is, in order to increase the average particle size when there are a crushing step and a classification step,
Three methods are conceivable: a method of coarsely crushing in the crushing step, a method of removing small particle sizes from the product in the classifying step, and a method of adding large particle size to the product in the classifying step. But,
In these three methods, the ratio of the amount of powder obtained as a product to the amount of powder supplied to the classification step (yield, yield) is completely different. In addition, in the usual powder production process, only a distribution with a certain spread is obtained in the pulverization process, and in the classification process as well, it is not sharply separated by the particle size called the classification point, and a certain spread is obtained. Be classified with. In addition, the process variables of the crushing process and the classification process are intricately entangled with each other, affecting the particle size distribution and yield of the product. Therefore, the method of controlling the difference directly with the process variable in order to make the particle size distribution of the product into a desired particle size distribution is the case that the influence of the process variable on the particle size distribution is not orthogonal to each other. It doesn't always work. In other words, in order to perform operation management, it was necessary to analyze the actual results and experience, and it was necessary to determine the set value of the process variable based on intuition and experience for a completely new product type. Due to these factors, it has been extremely difficult to predict and control the influence of the operation of process variables on the particle size distribution of products and the yield in the classification process.

[0004]

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and based on the numerical values of the particle size distribution before and after classification in the classification step, commercialization before and after classification according to a certain particle size The ratio (partial classification ratio) is calculated, and this partial classification ratio is defined as the partial classification ratio model that shows the performance of the process. The changes in the partial classification ratio model due to the operation of process variables are stored as a database, and the total process It is characterized by controlling while predicting the change in particle size distribution due to the operation of variables and the yield in the classification process.

That is, the gist of the present invention is a control method of a powder production method having a pulverizing and classifying step, wherein a partial classification ratio model is obtained from the particle size distributions before and after classification, and the partial classification set arbitrarily. By changing the parameters in the proportion model to determine the parameters in the partial classification proportion model that achieve the desired particle size distribution and yield after classification, and through a database that associates the parameters in the partial classification proportion model with the process variables. According to the control method of the powder production method, which is characterized by obtaining a process variable from the parameters in the partial classification ratio model determined by the method, and

A partial classification ratio model is obtained from the crushing part, the classification part, a measuring part for measuring the particle size distribution before and after the classification, and the particle size distribution before and after the classification, and the parameters in the partial classification ratio model are set arbitrarily. By changing
By determining the parameters in the partial classification ratio model that achieve the desired particle size distribution and yield after classification, and determining the parameters in the partial classification ratio model and the process variables, the determined partial classification ratio model is determined. The present invention resides in a powder production apparatus having a calculation unit for obtaining a process variable from an internal parameter.

With such a configuration, the influence of the process variable is recognized as parameters independent of each other in the partial classification ratio model (parameters in the partial classification ratio model),
Since the classification results such as the particle size distribution and yield of the product can be recalculated by using these independent parameters, an invariant control method can be constructed. Therefore, since the influence of the process variables on the parameters in the partial classification ratio model is considered to be peculiar to the device, it is possible to accumulate the invariant knowledge of the device, which is not dependent on the product type.

[0008]

In the control device according to the present invention, the particle size distribution data before classification, which is measured before the classification process after the crushing process, and the particle size distribution data after classification, which is the object of control, are input. A partial classification ratio model is created by calculating the partial classification ratio from the pre-classification particle size distribution data and the post-classification particle size distribution data, and the pre-classification particle size distribution data and the post-classification particle size distribution data are matched by the partial classification ratio model and the classification is performed. Product yield in the process is required. Through the database representing the relationship between the partial classification ratio model parameter and the process variable, the process variable that achieves the desired particle size distribution and yield from the current state is obtained, and the manipulated variable of the process variable is output. When the manipulated variable of the process variable is presented, the particle size distribution after the manipulation is output.

[0009]

EXAMPLES In the present invention, a jet mill, a rotary impact crusher or the like can be used as the crusher, and a known crusher or classifier such as an airflow classifier can be used as the classifier. FIG. 1 is an explanatory view of one using an impact type crusher and a multi-product simultaneous classifier as an example of using the control device of the present invention. As the process variable of the impact type crusher, the rotation speed of the rotor is used, and in the multi-product simultaneous classification machine, the gas flow rate to the fine powder outlet and the gas flow rate to the coarse powder outlet are used. In the impact type crusher, even the clearance between the rotor and the stator is crushed, so that the crushed powder becomes finer by increasing the "rotor speed". The multi-product simultaneous classifier classifies the powder into fine powder, medium powder, and coarse powder by the air flow, and the product is the medium powder. Also, it can be assumed that there is no interference such that fine powder and coarse powder are mixed in due to the structure. Therefore, by increasing the "gas flow rate to the fine powder outlet", the flow of powder is moved from medium powder to fine powder.
Lower the proportion of medium powder and increase the proportion of fine powder. At the same time, the flow rate of the powder is moved from the intermediate powder to the coarse powder by increasing the “flow rate of gas to the coarse powder outlet”, so that the ratio of the intermediate powder is decreased and the ratio of the coarse powder is increased. Here, "average particle size of pulverized product", "volume 50% classification point between fine powder and medium powder" and "volume 50% classification point between medium powder and coarse powder" are set as parameters in the partial classification ratio model, and Control in combination. FIG.
Shows the volume particle size distribution of the powder after pulverization, that is, before classification and the intermediate powder after classification, that is, the volume particle size distribution of the product, and the numerical distribution of the partial classification ratio model, and is an explanation of the parameters within the partial classification ratio model. The channel on the horizontal axis of the distribution graph indicates the size of the particle size, and by convention, the LOG scale of the particle size is often used. Although the particle size distribution measuring instrument used this time generates only 16 channels, it is possible to form a smooth curve as shown in FIG. 2 by performing appropriate smoothing. Specifically, a smooth curve can be realized by further subdividing each channel, assuming a numerical value, and repeating the moving average so as to be smooth while preserving the volume in each channel. As shown in FIG. 2, x represents a channel representing the particle size, and the volume particle size distribution B before classification is
(X), the volume particle size distribution M (x) of the product, and the numerical distribution F (x) of the partial classification ratio model. Each of these has the following properties.

[0010]

[Equation 1] From normalization ∫B (x) dx = 100% (integration range is the whole section) From normalization ∫M (x) dx = 100% (integration range is the whole section) Fine powder and coarse powder are mixed in From the above, max (F (x)) = 1 Further, from the mass conservation law of the partial classification ratio model min (F (x)) = 0 From the above, the partial classification ratio model is as follows. M (x) × a = F (x) × B (x) a: Product yield

By using the above relationship, a and F (x) can be obtained from the measured B (x) and M (x). More specifically, the maximum value of the mountain-shaped function G (x) obtained by dividing M (x) by B (x) by the same x is h.
Then, the reciprocal of h is the product yield a, and F (x) is obtained by dividing G (x) by h and normalizing the maximum value to 1.0. Change the shape of the partial classification ratio model by moving the "average particle size of pulverized product", "fine powder-volume 50% classification point of medium powder""medium powder-volume 50% of coarse powder" of the parameters in the partial classification ratio model, B '(x) and F' (x) are calculated (classification result prediction unit in FIG. 1). More specifically, F (x)
Change from 0.0 to 1.0 on the left and right with 1.0 as the boundary and 1.
The change from 0 to 0.0 is FR (x), FL respectively
(X) It is defined as follows.

[0012]

(2) F (x) = FR (x) × FL (x)

Various relationships among the parameters in the partial classification ratio model can be considered, but here it is simply assumed as follows.

[0014]

F ′ (x) = FR (x + y1) × FL (x + y2) y1: Fine powder-medium powder volume 50% classification point change amount y2: Medium powder-coarse powder volume 50% classification point change amount

Various parameters in the partial classification ratio model of the impact type crusher can be considered, but here it is simply assumed as follows.

[0016]

## EQU4 ## B '(x) = B (x + y3) y3: change in average particle size of crushed product

Then, the particle size distribution M '(x) of the product and the yield a'of the product are calculated using the formula of the partial classification model, and the optimum (desired) partial classification ratio model internal parameter is determined. Specifically, it is calculated by the following formula.

[0018]

## EQU5 ## N (x) = B '(x) × F' (x) a '= ∫N (x) dx (the integration interval is the entire interval) M' (x) = N (x) / a '

Each partial classification ratio model internal parameter corresponds to a process variable via a database. As the simplest example of associating the parameters in the partial classification ratio model with the process variables in the database, y1, y2, and y3 are multiplied by a gain function to obtain a “fine powder outlet gas flow rate” and a “coarse powder outlet gas flow rate”, respectively.
This can be achieved by feeding back the "rotor speed" in a speed type. Of course, methods other than the above can be considered for associating the parameters in the partial classification ratio model with the process variables.

[0020]

As described above, according to the present invention, a partial classification ratio model is created from the measured value of the particle size distribution before classification and the particle size distribution of the product, and the product yield and the product are calculated using the partial classification ratio model. By predicting and optimizing the particle size distribution of, it is possible to construct an invariant control method that is not affected by the type of powder.

[Brief description of drawings]

FIG. 1 is a diagram showing a control system of an impact type crusher and a multi-product simultaneous classifier for carrying out the present invention.

FIG. 2 is an explanatory diagram of a partial classification ratio model and parameters in the partial classification ratio model. Horizontal axis Particle size Vertical axis Volume ratio

Claims (2)

[Claims] 1. A method for controlling a powder production method comprising a pulverization and classification step, wherein a partial classification ratio model is obtained from particle size distributions before and after classification, and a parameter in the partial classification ratio model set arbitrarily is changed. Then, the parameters within the partial classification ratio model that achieve the desired particle size distribution and yield after classification are determined, and the determined partial parameters are passed through the database that associates the parameters within the partial classification ratio model with the process variables. A method for controlling a particle production method, characterized in that a process variable is obtained from a parameter in a classification ratio model. 2. A partial classification ratio model obtained by obtaining a partial classification ratio model from the crushing part, the classification part, a measurement part for measuring the particle size distribution before and after classification, and the particle size distribution before and after classification. By changing the internal parameter to determine the partial classification ratio model internal parameter that achieves the desired particle size distribution and yield after classification, and through the database that associates the partial classification ratio model internal parameter with the process variable, A powder production apparatus comprising a calculation unit for obtaining a process variable from the determined parameters in the partial classification ratio model.