JPH05262590A - System for controlling production of compound fertilizer - Google Patents

Abstract

PURPOSE:To improve yield and to save labor by making particle diameters of granulated compound fertilizer close to target particle diameters by intermittent photographed image of the granulated compound fertilizer and carrying out specific control in a stable state and in unstable state. CONSTITUTION:A raw material 1 is mixed with a returned raw material 2 by a blender 3 and thrown into a granulator 10. In the granulator 10, the blended raw materials are mixed with granulating water 4, sulfuric acid 5, ammonia 6 and phosphoric acid 7 rolled and granulated. The amounts of the granulating water 4, the sulfuric acid 5, the ammonia 6 and the phosphoric acid 7 indicated in product specifications are fed in the vicinity of the granulator 10. The production of compound fertilizer is controlled by using a collected table called particle size diagnosing S/H matrix 100 and integrating the table into a computer. A skilled specialist of granulation grasps the state of particles sent from the granulator 10 by observation, sieve and touch, reasons by using various knowledge of granulation and rules by experience through the granulated state during the operation and experience for many years and solves problems.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a chemical fertilizer production control system, and more particularly to an automatic control of production by a control computer incorporating an optimum control algorithm and the judgment and operation of a skilled granulation expert. ..

[0002]

2. Description of the Related Art Conventionally, when performing automatic control in the production of chemical fertilizers, the flow of the produced granular fertilizer is captured as a still image, the weight particle size distribution is calculated from this still image, and the weight ratio of the product size, Calculate the weight ratio of fine powder and the weight ratio of coarse particles, and recognize and adjust the granulated state of the granular fertilizer by comparing the above weight ratios with the weight ratios when ideally manufactured. Therefore, the granulated state was controlled to approach the ideal value. In addition, this control uses fuzzy control based on a rule based on the experience knowledge of the operator. For example, when the granulation state is recognized as "a little small", the water content is increased by 0.1% of the weight, and the granulation state is "very small". When it is recognized that the water content is increased by 0.2% of the weight, the granulation state is controlled only by adjusting the water content based on fuzzy expressions such as "a little small" and "quite small". It was

[0003]

In the conventional automatic control method, for example, when the weight ratio is biased and it is attempted to be corrected, even if the state is close to the target because the past transition is not taken into consideration, I sometimes overshooted the target by performing an operation to bring it closer to the target. Further, since the control is performed by the weight ratio, even if the weight ratio does not change, the state of the particles may move in a direction deviating from the target, and this cannot be detected early.

When the control is unstable and cannot be controlled only by ordinary control using water or steam, a stable state can be restored unless control other than water is performed according to the empirical rule that granulation experts usually use. There was a drawback that I could not.

[0005]

DISCLOSURE OF THE INVENTION The present invention has an object to achieve the retention of the granulation portion to the same extent as the operation of a skilled person even when the granulation is automated by solving the above-mentioned problems.

In a production control system for producing chemical fertilizer by automatic control, the granulated chemical fertilizer is intermittently photographed on a belt conveyor at the exit of the granulator, and the center of the particle size distribution is image-processed from the photographed image. To determine the central particle size. At this time, if the particle size predicted from the transition of the measured series of central particle sizes is different from the target particle size, one or several flow rates of water, steam, sulfuric acid, phosphoric acid and ammonia are controlled to control the target particle size. Move the product closer to the diameter. For the control in the normal stable state, the optimum control based on the mathematical model of the granulation process is used for control, and when the control is in an abnormal state and is not stable, the operation method of a skilled granulation expert and It is characterized by automatic control using a particle size diagnosis matrix table incorporating granulation knowledge.

[0007]

An embodiment of the present invention will now be described in detail with reference to the drawings.

Referring to FIG. 1, a chemical fertilizer production control system according to this embodiment comprises a chemical fertilizer raw material 1 and a return raw material 2.
, Mixer 3, granulated water (steam) 4, sulfuric acid 5, ammonia 6, phosphoric acid 7, electromagnetic flow meter 8, electromagnetic valve 9
, Granulator 10, belt conveyor 11, particle size distribution sensor 12, pH sensor 13, and temperature sensor 14
, Flow rate control unit 15, flow rate measurement unit 16, granulation expert system 17, dryer 18, cooler 19
And a sorting machine 20, a crusher 21, and a product 22. The chemical fertilizer targeted by the present invention has a particle size of 0.
It is a granular material of about 7 to 7.0 mm, which is formed by mixing raw materials containing lime superphosphate and potassium chloride, and wet chemical phosphoric acid and ammonia or ammonium phosphorus, ammonium sulfate, ammonium chloride, urea, potassium chloride and sulfuric acid. There is advanced chemical compounding that is made by mixing raw materials containing potassium.

FIG. 2 shows a granularity diagnosis S / H matrix 100.
This is an example of the event name part 101 and the hypothesis part 10
2, a condition unit 103, a certainty factor unit 104, and a certainty factor symbol unit 105.

In the condition unit 103, the T condition, the AND condition and the nOR condition are "if each event occurs, change the likelihood of the hypothesis", "if event 1 (occurs, does not occur) and , Event 2 (occurs,
If it does not occur) and ..., then the certainty of the hypothesis is the specified certainty factor "and" if event 1 (occurs, does not occur), event 2 (occurs, does not occur),
If n or more of the above conditions are satisfied,
"Set the certainty of the hypothesis to the specified certainty".

In the certainty factor unit 104, the numerical value represents the certainty factor (-1.0 to 1.0).

In the certainty factor symbol section 105, the certainty factor symbol differs depending on the condition.

That is, when the condition unit 103 is T, ◎, ○, △, ▲, ●, ★, × (1.0 ≧ ◎>○> △
＞ ▲ ＞ ● ＞ ★ ＞ ≧ × = -1.0)
And nOR are represented by T and F (T = “occurs”, F = “does not occur”), respectively.

Next, the operation of this embodiment will be described with reference to FIGS.

Granulation of the chemical fertilizer is as follows: raw material (several types) 1
After mixing the return raw material 2 with the mixer 3, it is charged into the granulator 10. In the granulator 10, the raw materials after mixing and the granulating water (steam) 4, sulfuric acid 5, ammonia 6, and phosphoric acid 7 are supplied to perform rolling granulation. The amount of granulated water 4, sulfuric acid 5, ammonia 6 and phosphoric acid 7 is added near the entrance of the granulator 10 as instructed in the product design document, but this varies greatly depending on the production amount of the brand and the state of the raw materials. To do. The chemical fertilizer granulated by the granulator 10 is conveyed to the dryer 18 by the belt conveyor 11 and dried, and then cooled by the cooler 19, and is sorted by the sorter 20 into a product having a particle size within the standard and other products. Be sorted. For defective products with non-standard particle size, after crushing large particles with the crusher 21,
It becomes the return raw material 2 and is reused as a part of the chemical fertilizer raw material. The product 22 having the particle diameter within the standard is packed in a bag. In the present invention, in the production of the above-mentioned chemical fertilizer, the particle size distribution is measured from a static image, the average particle size (center particle size) is compared with the target particle size, and if there is a difference, it is normally controlled automatically. The particle size is corrected, but if it becomes difficult to correct the particle size, the cause and the treatment method are derived from the particle size diagnosis S / H matrix 100, and granulated water (steam) 4, ammonia 5, sulfuric acid 6, phosphoric acid 7 Change the raw material input amount and return input amount according to the prescription.

The control method (control knowledge) required for controlling the production is compiled in a tabular form called a grain size diagnosis S / H matrix 100 and is used by being incorporated in a computer. Skilled granulation specialists have granulators 10
The state of the particles coming out from the
In order to take appropriate measures, we make inferences using various granulation knowledge and empirical rules obtained through many years of experience with the granulation state at that time, and solve the problem. In order to organize the problem solving methods and knowledge of such experts and make them easy to maintain, a table called a granularity diagnosis S / H matrix 100 is used. Grain state and grain size diagnosis S of particles emitted from the granulator 10
The particle states described in the / H matrix 100 are matched, the most likely hypothesis is derived from the matched symptom sequence, and control corresponding to the hypothesis is performed.

In order to grasp the state of particles moving from the granulator 10 on the belt conveyor 11, it is necessary to capture a still image from a camera and analyze it by image processing. In the factory to which the present embodiment is applied, there are only a limited number of places on the production line where real-time imaging is possible. Therefore, in order to capture a still image of the chemical fertilizer, a semi-finished product from the outlet of the granulator 10 to the dryer 18 It is optimal to provide the particle size distribution sensor 12 on the belt conveyor 11 that conveys the particles. The captured still image is converted into a binary image by a computer, and the particle size distribution is measured from this binary image. Furthermore, the particle size with the highest weight ratio (central particle size) is obtained from the particle size distribution, and comparison is made to see if it is within the target particle size range. When the grain state on the belt conveyor 1 is not stabilized by the usual automatic control by the optimal control, the symptom is searched for in the grain size diagnosis S / H matrix 100 to derive some hypotheses, and the most probable one is found. Take the procedure that corresponds to the higher hypothesis and perform the procedure similar to what a trained granulation expert would do. For example, when the state of particles on the belt conveyor 11 is such that the ratio of large particles and fine particles is high at the same time and the ratio of the target particle size is low, it is impossible to control with ordinary granulating water, but it is incorporated in a computer. By examining the particle size diagnosis S / H matrix 100, an intermediate hypothesis of “high product temperature” and “high pH” is obtained, and the most likely one of the hypotheses is selected and written in the matrix. Control according to procedures. When the particles are large, the causes are "excessive water content", "high granulated product temperature", "low pH", "abnormal ammonia sparger", and "abnormal water sparger". The particle size diagnosis S / H matrix 10 is based on the hypothesis that "there is a change in raw materials", "the amount of return is decreasing".
The treatment is carried out according to the most probable hypothesis, which is derived from 0. For example, in the case where the state of particles on the belt conveyor 11 is “fine particles”, “powdered sandy” is conspicuous, “large particles”, and “there are many large particles”, the hypothesis unit 102 of FIG. Of these, the “high pH” hypothesis is most likely. Therefore, the treatment is ammonia 6
The amount of sulfuric acid needs to be decreased or the amount of sulfuric acid 5 increased. However, by changing the amounts of ammonia 6 and sulfuric acid 5, the pH value detected by the pH sensor 13 may deviate from the standard, and the granulated product temperature detected by the temperature sensor 14 may change. Then, control is performed. Therefore, the compounding ratio of the chemical fertilizer raw material 1, the current flow rate of the return raw material 2, the ammonia 6, the sulfuric acid 5, the current set values of the granulation water 4,
The granulation expert system 17 calculates based on the temperature of the granulated product detected by the temperature sensor 14, the pH value conforming to the standard, and the quality component guarantee, and as a result, the solenoid valve 9 for ammonia 9 and sulfuric acid 5 is calculated. Adjust. When automatically controlling the grain state in the production of chemical fertilizers, the following items are targeted.

(1) Input amount of chemical fertilizer raw material (2) Input amount of return material (3) Granulation water, steam flow rate (4) Ammonia flow rate (5) Sulfuric acid flow rate (6) Phosphoric acid flow rate (7) Granulated product Temperature (8) pH value In order to make the state of the particles on the belt conveyor 11 at the exit of the granulator equal to that of granulation by a skilled operator, the particle size diagnosis S / H matrix based on the empirical knowledge of the expert. 100 is used on a computer to perform an optimum operation.

Usually, in the case of a skilled granulation expert, the grain condition is monitored about once every 10 to 15 minutes on average.
In the case of automatic control, it is performed once every 10 seconds.

[0020]

As described above, according to the present invention, even in the granulation process which is not stable in the conventional automatic control system, the computer is performed by using the particle size diagnosis S / H matrix incorporating the expert's granulation know-how. By controlling, it is possible to improve the yield and to save labor by reducing the intervention of the operator.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing an embodiment of the present invention.

FIG. 2 is an example of a particle size diagnostic S / H matrix used in this embodiment.

[Explanation of Codes] 1 Chemical fertilizer raw material 2 Return raw material 3 Mixer 4 Granulating water (steam) 5 Sulfuric acid 6 Ammonia 7 Phosphoric acid 8 Electromagnetic flow meter 9 Electromagnetic valve 10 Granulator 11 Belt conveyor 12 Particle size distribution sensor 13 pH sensor 14 Temperature sensor 15 Flow rate control unit 16 Flow rate measurement unit 17 Granulation expert system 18 Dryer 19 Cooler 20 Sorter 21 Crusher 22 Products

Claims (1)

[Claims] 1. A production control system for producing chemical fertilizer by automatic control, the granulated chemical fertilizer is intermittently photographed on a belt conveyor at the exit of the granulator, and the center of the particle size distribution is taken from the photographed image. When the measured particle size is different from the target particle size predicted from the transition of the measured central particle size, the target is adjusted by adjusting one or several flow rates of water, steam, sulfuric acid, phosphoric acid and ammonia. The particle size of the granules is controlled to be close to the particle size, and the optimum control is used for the control in the normal stable state. If the control is not stable, the operation method and granulation knowledge of a skilled granulation expert are incorporated. A chemical fertilizer production control system characterized by automatic control using a particle size diagnosis matrix table.