JP2962793B2 - Flow control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to a method in which a flow rate of an object to be weighed such as powder, granules, or a liquid derived from a storage tank such as a measuring tank or a storage tank is substantially constant. Operation control, for example, loss-in-weight type,
The present invention relates to a flow control device such as a belt type.

[Conventional technology]

First, a loss-in-weight type flow control device will be described as an example.

In FIG. 33, a weighing tank 1 serving as a storage tank is provided with a weighing device 2 such as a load cell, and an object to be weighed such as powder in the weighing tank 1 is led out from below by a screw feeder 3. The analog weighing signal from the weighing device 2 is
The signal is converted into a digital weighing signal by the A / D converter 4 and supplied to the controller 5 for the screw feeder 3. The controller 5 calculates the flow weight value and the cumulative weight value per unit time of the object to be weighed, which are derived from the weighing tank 1 by the screw feeder 3 based on the digital weighing signal, and calculates the calculated flow weight value. And the accumulated weight value are compared with the set target flow rate weight value and target accumulated weight value. The operation of the screw feeder 3 is controlled based on the comparison result such that the flow rate of the object to be weighed out of the weighing tank 1 becomes substantially constant.

Conventionally, as a control system of the flow control device as described above, the target flow weight value and the coefficient of the feedforward proportional element set as shown in FIG. An operation control amount is calculated from K _ff, and the operation control amount is
Specifically, a feedforward control system supplied to the screw feeder 3 is used.

[Problems to be solved by the invention]

Incidentally, in the feedforward control system, if the coefficient K _fb coefficient K _ff and feedback proportional element of the feedforward proportional element is properly configured, the objects to be weighed as shown in FIG. 34 Both the flow weight value and the cumulative weight value per unit time are controlled according to the target flow weight value and the target cumulative weight value. However, for example, when the coefficient K _ff of the feedforward proportional element is not properly set, the flow weight value per unit time is equal to the target flow weight value as shown in FIGS. 35 and 36. However, an offset occurs in the accumulated weight value.In order to eliminate the offset, the coefficient K _fb of the feedback proportional element is set to be large, as shown in FIG. 37. Is the same as the target cumulative weight value, but the flow rate weight value fluctuates greatly and cannot be said to be in an appropriate control state.

The state where the setting of the coefficient _Kff of the feedforward proportional element is not appropriate occurs under the following conditions.

1. Change or incorrect setting of bulk density of weighing object from setting time 2. Change of flow characteristic of weighing object 3. Change of feeder efficiency Therefore, in the case of feedforward control system, control characteristics and control accuracy are maintained To do this, for each change in bulk density, flow characteristics, and feeder efficiency of the object to be weighed,
There is a problem that the coefficient K _ff of the feedforward proportional element must be reset.

An object of the present invention is to solve such a problem without using a conventional feedforward proportional element by using a fuzzy control system as a flow rate control system for an object to be weighed, and using a conventional feedforward proportional element. In the case where the coefficient is used, for example, a flow control device capable of maintaining good control characteristics and control accuracy without resetting the coefficient of the feedforward proportional element for each change in the bulk density and flow characteristics of the object to be weighed. To provide.

[Means for solving the problem]

In order to achieve the above-described object, the flow control device according to the present invention has a structural feature as shown in FIG. A flow controller for operating and controlling the feeder so that the flow rate is substantially constant, comprising: (a) a target flow weight value, a target cumulative weight value per set unit time, and a weight value based on a signal obtained from the measuring device. A variable value calculation unit (1) for calculating the deviation flow weight value, the variation amount of the deviation flow weight value, and the deviation cumulative weight value by using (1), (b) the deviation flow weight value and the deviation flow as the control rule antecedent variables. A first fuzzy inference unit (2) for obtaining a first fuzzy inference value by a fuzzy inference operation using the change amount of the weight value; and (c) fuzzy inference using the deviation cumulative weight value as a control rule antecedent variable. Inference performance A second fuzzy inference unit for obtaining a second fuzzy inference value by calculation, and (d) based on addition of the first and second fuzzy inference values by the first and second fuzzy inference values at a predetermined ratio. An operation control amount calculation unit (4) for obtaining an operation control amount, or (d) an absolute value of a first fuzzy inference value among the first and second fuzzy inference values by the first and second fuzzy inference units. Fuzzy control means having an operation control amount calculating section (4) 'for obtaining an operation control amount based on a larger selection by comparing the second control value with an absolute value obtained by multiplying a second fuzzy inference value by a predetermined coefficient. It is.

In addition, in order to obtain the second fuzzy inference value in the second fuzzy inference unit (2), in addition to the biased cumulative weight value from the target cumulative weight value and a change in the deviation cumulative weight value as a control rule antecedent variable. It may be obtained by a fuzzy inference operation using the quantity.

[Action]

FIG. 6 to FIG. 6 to FIG. 6 to FIG. 6 which are used in the description of an embodiment in which the bulk densities of the objects to be weighed are different when obtaining the operation control amount based on the addition of the first and second fuzzy inference values at a predetermined ratio in the present invention. 8, the flow weight value and the cumulative weight value per unit time converge to the target flow weight value and the target cumulative weight value.

Similarly, of the latter first and second fuzzy inference values, the absolute value of the first fuzzy inference value is compared with the absolute value obtained by multiplying the second fuzzy inference value by a predetermined coefficient. In the case where the operation control amount is obtained based on the selection of the weighing object, the bulk density of each of the objects to be weighed is different from that of the eighteenth embodiment.
Explaining with reference to the control characteristic diagrams in FIGS. 20 to 20, all of the values converge to the target flow rate weight value and the target cumulative weight value.

〔The invention's effect〕

As described above, according to the first and second aspects of the present invention, the flow rate and the cumulative weight value are determined by the operation control amount obtained using the first fuzzy inference value regarding the flow rate and the second fuzzy inference value regarding the cumulative weight value. It can be made to converge to the target value based on it. Therefore, it is possible to maintain good control characteristics and control accuracy without resetting the parameters such as the feedforward proportional element for each change in the bulk density and flow characteristics of the object to be weighed.

Further, in the invention according to claim 3, the coefficient of the feedforward proportional element used in the conventional feedforward control system is input to the variable value calculation unit, and is used as the operation control amount at the start of operation obtained using this coefficient. By using the initial value, the time required for both the flow rate and the cumulative weight value to converge to the target value can be reduced as compared with the case where the initial value is not used, that is, the effect of improving the start-up characteristics can be achieved. . The coefficient of the previous feedforward proportional element is used only for calculating the initial value.

Further, in the invention according to claim 4, the value of the feeder system proportional element is calculated based on the operation control amount during operation and the target flow rate weight value, and the calculated value of the feeder system proportional element is out of the predetermined range. In the first and second cases, the operation control obtained by the operation control amount calculation unit is corrected by modifying the consequent variable of the control rule of the fuzzy inference unit to a variable value obtained by the value of the calculated feeder system proportional element. This has the effect of reducing the time required for the fuzzy set to converge on the amount of change in the amount.

〔Example〕

Next, a specific embodiment of the flow control device according to the present invention will be described with reference to the drawings.

FIG. 2 shows a controller for a screw feeder in a loss-in-weight type flow control device for drawing out an object to be weighed such as powder in a weighing tank (not shown) from below. Similarly, digital weighing signals digitally converted by the A / D converter from a weighing device provided in a weighing tank (not shown) are sequentially provided to the variable value calculation unit 21 at a predetermined sampling interval. In the variable value calculator 21, the following calculation is performed based on these digital weighing signals.

Per unit time of the flow rate by weight value X (g / s): X = (W n -W n-1) / t W n (g); weight value based on the digital weight signals by the current sampling W _n-1 (g ); Weight value based on digital weighing signal from previous sampling t (s); Sampling interval Deviation flow weight value E _x (g / s): E _x = Q-X (1) Q (g / s); Unit time per target flow rate weight value deviation rate weight value E _X (g / s) of the sampling interval t (s) for each amount of change _{ΔE x (g / s):} ΔE x = E xn -E xn-1 (2) E _xn ; Deviation flow weight value calculated at the time of the current sampling E _xn-1 ; Deviation flow weight value calculated at the time of the previous sampling On the other hand, the following calculation is also performed to adopt the cumulative weight control as the flow control .

Target cumulative weight value after T (s) from the start of flow rate control R (g): R = Q × T Deviation cumulative weight value E _Y (g): E _Y = R−W _n (3) Wn (g); flow rate cumulative weight value after nt (s) from the control start Next, (1), (2) the error to flow weight values E _X obtained by the equation (g / s) and the deviation rate weight value E _X (g / s) The change amount ΔE _X (g / s) at each sampling interval t (s) is given to the first fuzzy calculation unit 22A, and the deviation cumulative weight value E _Y (g) obtained by the equation (3) is 2 fuzzy operation section 22B. In the first fuzzy operation unit 22A, fuzzy operation is performed based on the control rules constituting the first control rule from the first control rule unit 23A. Incidentally, the first control rule deviation rate weight value E _X
(G / s) and the variation ΔE _X (g / s) at each sampling interval t (s) of the deviation flow rate weight value E _X (g / s) as the antecedent, and the operation control amount variation dU (v) Is the consequent part. As a result of the inference by the fuzzy operation, the first
First as a fuzzy inference value of the operation control quantity variation dU ₁ is obtained. Further, in the second fuzzy calculation section 22B, fuzzy calculation is performed based on the control rules constituting the second control rule from the second control rule section 23B. Note that the second control rule is based on the deviation cumulative weight value E
Let _Y (g) be the antecedent and let the operation control variable change dU (v) be the consequent. As an inference result by this fuzzy operation, a second operation control amount change amount dU ₂ is obtained as a _second fuzzy inference value for the operation control amount.

Next, the fuzzy inference method used in the present embodiment will be specifically described.

As a fuzzy inference method, the MAX-MIN method of the Mumdani method based on the composition rule of fuzzy relations is used,
The defuzzification uses the centroid method, and the inference formula, in other words, the antecedent and consequent parts of the control rule both use the triangle shown in FIG. 3 as the membership function of the fuzzy variable. The fuzzy level in FIG. 3 has the following meaning.

 NB; Negative and large. NM; negative and medium.

 NS; negative and small. ZO; zero.

 PS; positive and small. PM; positive and moderate.

 PB; positive and large.

Note that S _max and S _min in FIG. 3 are fuzzy labels NB
= 1 gives PB = 1 a fuzzy variable value, the deviation rate weight value E _X (g / s), the amount of change in each sampling interval t (s) of the deviation rate weight value E _X (g / s) In a fuzzy set of ΔE _X (g / s), the cumulative deviation weight value E _Y (g), and the operation control amount change amount dU (v), they are expressed as follows.

Deviation rate weight value _{E X (g / s):} S max = K eq × Q max + Q O S min = -S max deviation rate weight value E _X (g / s) of the sampling interval t
Change amount ΔE _X (g / s) for each (s): S _max = K _dep × Q _max + Q ′ _O S _min = −S _max Deviation cumulative weight value E _Y (g): S _max = K _ew × Q _max + E _YO S _min = −S _max Q (g / s); target flow rate weight per unit time Q _max (g / min); maximum target flow rate weight per unit time Q _O , Q ' _O (g / min); control in the sense of raising Elevated flow weight per unit time set according to the system E _YO (g / min); Cumulative weight per unit time deviation set according to the control system in the sense of elevating C _O ; Constant K _eq, K _deq; control system of the mechanism characteristics and the coefficient determined by the characteristics of the objects to be weighed K _ew (min); time K _cmp; operation control quantity weight coefficient Incidentally, deviation rate weight value E _X (g / s) and deviation rate weight value E _X (g / s) of the sampling interval t (s) for each amount of change Delta] E _X a (g / s) and antecedent, the operation control quantity variation dU
As an example of the first control rule having (v) as a consequent part,
It is as follows.

_{No.1; if E X = PB and} ΔE X = ZO then dU = PB No.2; if E X = ZO and ΔE X = NB then dU = NB No.3; if E X = NB and ΔE X = ZO then dU = NB No.4; if E _X = ZO and ΔE _X = PB then dU = PB No.5; if E _X = PM and ΔE _X = ZO then dU = PM No.6; if E _X = ZO and ΔE _X = HM then dU = NM No.7; if E _X = NM and ΔE _X = ZO then dU = NM No.8; if E _X = ZO and ΔE _X = PM then dU = PM No.9; if E _X = PS and ΔE _X = ZO then dU = PS No. 10; if E _X = ZO and ΔE _X = NS then dU = NS No. 11; if E _X = NS and ΔE _X = ZO then dU = NS No. _{12; if E X = ZO and} ΔE X = PS then dU = PS No.13; if E X = ZO and ΔE X = ZO then dU = ZO When indicating this first control rule table, the following table 1 It is as follows.

Further, an example of the second control rule in which the deviation accumulated weight value E _Y (g) is the antecedent and the operation control amount change dU (v) is the consequent is as follows.

No.1; if E _Y = NB then dU = NB No.2; if E _Y = NM then dU = NM No.3; if E _Y = NS then dU = NS No.4; if E _Y = ZO then dU = ZO No.5; if E _Y = PS then dU = PS No.6; if E _Y = PM then dU = PM No.7; if E _Y = PB then dU = PB This is shown in Table 2 below.

Next, an operation method using the above-described membership function according to the first and second control rules will be described.

First, regarding the ∧ (min) operation in the antecedent part of the control rule, taking the No. 1 control rule of the first control rule as an example,
To be described with reference to Figure 4, the deviation rate weight value E _X (g / s) obtained by the variable value calculation unit 21 and the deviation rate weight value E _X
(G / s) change ΔE _X (g at each sampling interval t (s)
/ s) to a deviation rate weight value E _X1 and variation Delta] E _X1. Fit is obtained for the fuzzy label PB from the deviation rate weight value E _X1, also fuzzy label ZO of variation Delta] E _X1
Is obtained. These obtained as adaptability omega ₁ for fuzzy label PB fuzzy sets two values the operation control amount of change smaller of fitness dU (v).

Similarly, for the control rules No. _{2 to} No. ₁₃ in the first control rule, the fitness levels ω _{2 to} ω ₁₃ are obtained.

On the other hand, the final conformity of each fuzzy label in the consequent part in the control rule is calculated by V (max), and the final conformity of each fuzzy label NB, NS, ZO, PS, PM, PB is represented by ω _NB , ω _NM ,
Assuming that ω _NS , ω _ZO , ω _PS , ω _PM , ω _PB , the following calculation is performed.

ω _NB = ω ₂ ∨ω ₃ ω _NM = ω ₆ ∨ω ₇ ω _NS = ω ₁₀ ∨ω ₁₁ ω _ZO = ω ₁₃ ω _PS = ω ₉ ∨ω ₁₂ ω _PM = ω ₅ ∨ω ₈ ω _PB = ω ₁ ∨ω _{4 A} membership function of the operation control amount change amount dU (v) obtained in this way is as shown in FIG.

Next, defuzzification of the fuzzy set of the operation control amount change amount dU (v) is performed by using the centroid method, and the first operation control amount change amount dU ₁ as the first fuzzy inference value is obtained by the following equation. Can be

B (u '): Membership function of the operation control amount change amount dU (v) The fuzzy calculation based on the control rules constituting the first control rule is completed by the above calculation method.

Next, control rules that constitute the second control rule will be described.

Taking the No. 1 control rule of the second control rule as an example, let the deviation cumulative weight value E _Y (g) obtained by the variable value calculator 21 be the deviation cumulative weight value E _Y1 . The degree of conformity to the fuzzy label NB can be obtained from the deviation cumulative weight value E _Y1, but in this No. 1 control rule, the antecedent part is composed only of the deviation cumulative weight value E _Y1 , so The degree of conformity ω ₁ of the fuzzy set of the operation control amount change amount dU (v) to the fuzzy label NB is obtained as it is.

Similarly, for the control rules No. 2 to No. 7 of the second control rule, the fitness levels ω _{2 to} ω ₇ are obtained, but the same fuzzy label is used in the consequent part of the _seven control rules. Since only one different one of the seven fuzzy labels appears without appearing, the final fitness of each fuzzy label NB, NM, NS, ZO, PS, PM, PB is defined as ω _NB , ω _NM , ω _NS , ω _ZO , ω _PS , ω _PM , ω _PB are as follows.

ω _NB = ω ₁ ω _NM = ω ₂ ω _NS = ω ₃ ω _ZO = ω ₄ ω _PS = ω ₅ ω _PM = ω ₆ ω _PB = ω _{7 The} operation control amount change amount dU for the second control rule
The defuzzification of the fuzzy set of (v) is the same as that described for the first control rule, and the second operation control amount change amount dU ₂ as the second fuzzy inference value is obtained by the following equation. .

By the way, these first and second operation control amount change amounts
dU ₁ and dU ₂ are given to the operation control amount calculation unit 24,
In the operation control amount calculation unit 24, the final operation control amount change amount dU (v) is calculated by the following equation.

dU = dU ₁ + K _C × dU ₂ Note that K _C (≧ 0) is a deviation cumulative weight correction coefficient that determines the addition ratio of the first and second operation control amount change amounts dU ₁ and dU ₂ .

Based on this final operation control amount change amount dU (v),
This operation control amount U _n that is calculated by the following equation is applied to the screw feeder.

U _n = U _n-1 + dU U _n-1 ; Previous operation control amount Next, the setting of the initial value U ₀ at the start of operation for the operation control amount U (v) (initial value setting function) and the feeder system proportional element The correction of K (self-adjustment function) will be described.

The initial value U ₀ at the start of operation of setting the initial value U ₀ (the initial value setting function) operation control quantity U (v), calculated in the variable value calculation section 21 by the following equation, shown by dotted lines in FIG. 2 This is given to the operation control amount calculation unit 24 as described above.

Q (g / s): Target flow weight per unit time K _ff ; Coefficient of feed forward proportional element Correction of feeder proportional element K (self-adjustment function) Feeder proportional element Varies from moment to moment depending on the properties such as the bulk density and fluidity of the object to be weighed and the efficiency of the feeder system, and is expressed by the following equation.

Therefore, the operation control amount is calculated by the operation control amount
Each time U _n is calculated, the operation control quantity U _n are supplied to the variable calculation unit 21 from the operation control amount calculating section 24 as indicated by one-dot chain line in Figure 2. This variable calculator 21
In the feeder system proportional element K _n is used is calculated by Equation.

If the relationship of K _n > K _n-1 (1 + K _er ) or K _n <K _n-1 (1-K _er ) is satisfied with respect to the previous feeder system proportional element K _n−1 , the feeder system As proportional element K
Using 1/2 (K _n + K _{n -1} ), an equation that gives S _max and S _min of the fuzzy set of the operation control amount change amount dU (v) is defined as K = 1/2 (K _n + K _{n -1} ). , S _max and S _min are changed.

A control characteristic diagram obtained by an experiment using the above-described loss-in-weight type flow rate control device is as follows.

 Equipment used in the experiment Screw feeder Twin screw feeder SF-6 Maximum capacity 6l / h Maximum rotation speed 30r.pm Screw outer diameter φ20mm Screw inner diameter φ12mm Screw pitch 10mm (Drive motor 15W CD servo motor) Measuring instrument Load cell Rated 4kg (UH36 -4) The feeder proportional element K is given by the following equation.

According to the device specifications and the object to be weighed used in the experiment, the appropriate value of the feeder proportional element K is It is.

Further, the deviation cumulative weight value correction coefficient K _c defining a first and second operation control quantity variation dU _1, addition ratio of dU ₂ as 0.15, an experiment was conducted as follows.

Experiment 1 << When both initial value setting function and self-adjustment function are not provided >> When the initial value of the fuzzy set of the operation control variable change amount dU (v) and the feeder system proportional element K are different as shown in Table 3 below In order to obtain the respective control characteristic diagrams, the control characteristic diagrams as shown in FIGS. 6 to 8 were obtained.

In any of the control characteristic diagrams shown in FIGS. 6 to 8, both the flow weight value and the cumulative weight value per unit time converge to the target flow weight value and the target cumulative weight value. This shows the effectiveness of obtaining the operation control amount based on the addition of the operation control amount change amounts dU ₁ and dU ₂ .

Experiment 2 << In the case of having only the initial value setting function >> As shown in Table 3 of Experiment 1, the operation control amount change amount
When the initial value of the fuzzy set of dU (v) and the feeder proportional element K are different, each control characteristic diagram is obtained by setting the coefficient _Kff of the feedforward proportional element to 0.5, as shown in FIG. Control characteristic diagrams as shown in FIGS. 11 to 11 were obtained.

The shift control characteristic diagrams shown in FIG. 9 to FIG. 11 are also the same as those shown in FIG. 6 to FIG.
It can be seen that the rise characteristic is improved as compared with the control characteristic diagram shown in the figure.

Experiment 3 << In the case of having only the self-adjustment function >> Each control characteristic diagram when the initial value of the fuzzy set of the operation control amount change amount dU (v) and the initial value of the feeder proportional element K are different as shown in Table 4 below. In order to obtain, control characteristic diagrams as shown in FIGS. 12 to 14 were obtained.

In any of the control characteristic diagrams shown in FIGS. 12 to 14, the fuzzy set of the operation control amount change amount dU (v) is close to -0.19 to 0.19 (v) by the convergence time shown in Table 4. Has converged.

Experiment 4 << When both the initial setting function and the self-adjustment function are provided >> As shown in Table 4 of Experiment 3, the operation control amount change amount dU
In the case where the initial value of the fuzzy set and the initial value of the feeder proportional element K in (v) are different, each control characteristic diagram is obtained by setting the coefficient K _ff of the feed forward proportional element to 0.5, as shown in FIGS. As a result, control characteristic diagrams as shown in FIGS. 15 to 17 were obtained.

Each of the control characteristic diagrams shown in FIGS. 15 to 17 has an operation control amount change amount dU in about 30 seconds after rising.
The fuzzy set of (v) converges around -0.18 to 0.18 (v).

Also, it can be seen that the rise is quicker and the flow rate weight value is less fluctuated than the control characteristic diagrams of FIGS. 9 to 11 of Experiment 2.

If the foregoing experimental 1 to 4, the final operation control quantity changes first and second operation control quantity variation dU _1, the dU ₂ with at deviation cumulative weight value correction coefficient K _c is added at a predetermined ratio Quantity dU
(V) was obtained. However, the final operation control amount change amount dU (v) can also be obtained by calculating with the following equation.

dU = sign (max (dU ₁ , K _c × dU ₂ )) × max (dU ₁ , K _c × d _U2 ) where max (a, b) is a function representing the larger value of a and b absolute values , Sign (c) is a function representing the sign of c (a, b, c are arbitrary values).

In other words, the first operation the absolute value of the controlled variable change amount dU _1, second difference cumulative weight value correction coefficient K _c is multiplied operation control quantity variation dU of absolute value larger becomes better of the _two It is to obtain the final operation control amount change amount dU (v) based on the selection.

Based on the selection of the larger one, the experiment in which the final operation control amount change amount dU (v) is made to correspond to Experiments 1 to 4, and the same experiment is performed with the other conditions being the same.
The control characteristic diagrams of FIGS. 18 to 29 corresponding to the respective FIGS. 6 to 17 were obtained. These control characteristic diagrams are the same as those in the case of obtaining the final operation control amount change amount dU (v) based on the addition of the first and second operation control amount change amounts dU ₁ and dU _{2 at} a predetermined ratio. The characteristic diagram is shown. Therefore,
A first operation the absolute value of the controlled variable change amount dU _1, the second operation control quantity variation dU deviation cumulative weight value correction coefficient K _c is multiplied ₂
The effectiveness of obtaining the operation control amount based on the selection of the larger one of the absolute values of the above is shown.

In this embodiment, the fuzzy label shown in FIG.
NB = 1, PB = 1 is a fuzzy variable value which gives the S _max is, the deviation rate weight value E _X (g / s), each error to flow weight values E _X (g / s) of the sampling interval t (s) In the fuzzy set of the change amount ΔE _X (g / s) and the deviation cumulative weight value E _Y (g), they are expressed as follows.

Deviation rate weight value _{E X (g / s):} S max = K eq × Q max + Q O error to flow weight values E _X (g / s) of the sampling interval t
(S) Change amount ΔE _X (g / s): S _max = K _deq × Q _max + Q ′ _O Deviation cumulative weight value E _Y (g): S _max = K _eq × Q _max + E _YO However, fuzzy variables The value S _max is also expressed as:

Deviation rate weight value _{E X (g / s):} S max × Q max + K eq '+ Q + Q O error to flow weight values E _X (g / s) of the sampling interval t _(s) for each amount of change ΔE _X (g / s _{): S max = K deq ×} Q max + K deq '+ Q + Q O deviation cumulative weight value _{E Y (g): S max} = K ew × Q max + K ew' × Q + E YO K eq ', K deq'; K eq, A _coefficient K _ew ′ (min) determined by the mechanical characteristics of the control system and the characteristics of the object to be weighed similar to K _deq ; the same time as K _{ew In the} present embodiment, the second control rule is set to the deviation cumulative weight value E _Y (g) is the antecedent part, and the operation control amount change amount dU (v) is the consequent part. As in the first control rule, the deviation cumulative weight value E _Y (g) is as follows. And cumulative deviation weight
E _Y (g) Amount of change ΔE _Y (g) for each sampling interval t _(s )
May be set as the antecedent part, and the operation control amount change amount dU (v) may be set as the consequent part.

Variation ΔE _Y (g / s) of deviation cumulative weight value E _Y (g) at each sampling interval t _(s ): ΔE _Y = E _Yn −E _Yn-1 E _Yn ; Accumulation of deviation calculated at the time of the current sampling Weight value E _Yn-1 ; Deviation cumulative weight value calculated at the time of the previous sampling The fuzzy variable values S _max and S _min corresponding to the fuzzy labels NB = 1 and PB = 1 in FIG. The fuzzy set of the change amount ΔE _Y (g) at each sampling interval t _{(s) of the} weight value E _Y (g) and the deviation cumulative weight value E _Y (g) is expressed as follows.

Deviation cumulative weight E _Y (g): S _max = K _ew × Q _max + K _ew ′ × Q + E _YO S _min = −S _max Deviation cumulative weight E _Y (g) at each sampling interval t _(s) ΔE _Y (g): S _max = K _dew × Q _max + K _dew ′ × Q + dE _YO S _min = −S _max dE _YO (g / min): Set according to the control system in the same meaning as E _YO. The change amount of the cumulative deviation weight value per unit time K _dew , K _dew ′; the system number determined by the mechanical characteristics of the control system and the characteristics of the object to be weighed. The deviation cumulative weight value E _Y (g) and the deviation Variation ΔE of cumulative weight value E _Y (g) at each sampling interval t _(s)
Table 5 shows an example of the second control rule in which _Y (g) is the antecedent part.

Incidentally, with the time to obtain a second operation control quantity variation dU _2, except that carried out Ruishi the calculation to obtain the first operation control quantity variation dU ₁ is similar to the embodiment previously described, the operation control FIGS. 6 to 8 when the initial value of the fuzzy set of the amount of change dU (v) and the feeder system proportional element K are different as shown in Table 3, and neither the initial value setting function nor the self-adjustment function is provided. To obtain each control characteristic diagram corresponding to
Control characteristic diagrams as shown in FIGS. 30 to 32 were obtained.

In this embodiment, the flow rate control device of the loss-in-weight type has been described.However, the object to be weighed in the storage tank is derived from below the storage tank by a belt conveyor feeder, and the object to be weighed on the belt conveyor feeder is loaded into a load cell. It is needless to say that the present invention is also applied to a belt-type flow rate control device that operates and controls the traveling speed of the belt conveyor feeder so that the flow rate of the object to be measured is substantially constant by measuring with the measuring device.

[Brief description of the drawings]

FIG. 1 is a block diagram corresponding to the configuration of the present invention described in the claims, and FIGS. 2 to 17 are drawings for explaining a specific embodiment of the flow control device according to the present invention. 2 is a schematic diagram, FIGS. 3 to 5 are explanatory diagrams for explaining the fuzzy inference method, FIGS. 6 to 17 are control characteristic diagrams by experiments, and FIGS. 18 to 32. Fig. 33 is a control characteristic diagram by an experiment of a modified example, Figs. 33 to 38 are drawings for explaining a conventional example, Fig. 33 is a schematic diagram, Figs. 34 to 37 are control characteristic diagrams, Figs. FIG. 38 is a block diagram of the feedforward control system. 1 First fuzzy inference unit 2 Second fuzzy inference unit 3, 3 'Operation control amount calculation unit 21 Variable value calculation units 22A and 22B First and second fuzzy calculation units 23A, 23B: First and second control rule units 24: Operation control amount calculation unit

Continuation of the front page (56) References JP-A-2-35501 (JP, A) JP-A-2-93904 (JP, A) JP-A-1-173115 (JP, A) JP-A 1-253011 (JP) Kazuo Usui, "Batch Weighting and Feed Rate Control Systems", The International Journal of Automotive Industry & General Insurance, Kazuo Usui, JP-A-60-148601 (JP, U) , Powder Handling & Processing, Vol. 2, No. 2, Trans Tech Publications, June 1990, Germany. 129-135 Noboru Higuchi and two others, "Mixing and Mixing System Using Fuzzy Theory", Chemical Equipment, Industrial Research Institute, Ltd., Vol. 31, No. 5, May 1989, p. 47-51 Ryosuke Iida, one other, "Application of fuzzy control to chlorine injection process in front of water treatment plant", Keiso, Kogyo Gijutsusha, Vol. 31, No. 5, May 1988, p. 45-49 (58) Field surveyed (Int. Cl. ⁶ , DB name) G05B 13/00-13/02 G05D 7/00-7/06 JICST file (JOIS)

Claims (4)

(57) [Claims] 1. A flow control device for controlling the operation of a feeder so that the flow rate of an object to be weighed out of the reservoir provided with a weighing device by a feeder is substantially constant. A variable value calculating unit that calculates a deviation flow weight value, a change amount of the deviation flow weight value, and a deviation accumulation weight value using a target flow weight value, a target cumulative weight value, and a weight value based on a signal obtained from the measuring device; (B) a first fuzzy inference unit that obtains a first fuzzy inference value by fuzzy inference operation using the deviation flow weight value and the change amount of the deviation flow weight value as control variable antecedent variables; (c) control A second fuzzy inference unit for obtaining a second fuzzy inference value by fuzzy inference operation using the deviation cumulative weight value as a rule antecedent variable; and (d) a first and a second fuzzy inference. Flow control apparatus characterized by comprising a first and second fuzzy controller having an operation control amount calculation portion for obtaining an operation control amount based on the addition by a predetermined percentage of the fuzzy reasoning value by. 2. A flow rate control device for controlling the feeder so that the flow rate of an object to be weighed out from a storage layer provided with a weighing device by a feeder becomes substantially constant. A variable value calculating unit that calculates a deviation flow weight value, a change amount of the deviation flow weight value, and a deviation accumulation weight value using a target flow weight value, a target cumulative weight value, and a weight value based on a signal obtained from the measuring device; (B) a first fuzzy inference unit that obtains a first fuzzy inference value by fuzzy inference operation using the deviation flow weight value and the change amount of the deviation flow weight value as control variable antecedent variables; (c) control A second fuzzy inference unit for obtaining a second fuzzy inference value by fuzzy inference operation using the deviation cumulative weight value as a rule antecedent variable; and (d) a first and a second fuzzy inference. Comparing the absolute value of the first fuzzy inference value with the absolute value obtained by multiplying the second fuzzy inference value by a predetermined coefficient among the first and second fuzzy inference values, and performing an operation based on the selection of the larger one A flow control device comprising fuzzy control means having an operation control amount calculation unit for obtaining a control amount. 3. A coefficient of a feed-forward proportional element is input to the variable value calculation unit, an initial value as an operation amount at the start of operation is calculated using the coefficient, and the initial value is used as the operation control amount operation unit. The flow control device according to claim 1, wherein the output is output to a controller. 4. The variable value calculator calculates a value of a feeder proportional element based on an operation control amount during operation and a target flow weight value, and the calculated value of the feeder proportional element is out of a predetermined range. 4. The method according to claim 1, wherein the variable of the consequent part of the control rule of the first and second fuzzy inference units is corrected to a variable value obtained by the calculated value of the feeder proportional element. The flow control device according to any one of the above.