JPS58135423A - Automatic compensating method for automatic determining balance using computer - Google Patents

Abstract

PURPOSE:To improve both reliability and efficiency, by having prescribed step control by means of a computer and closing a gate by the cutoff weight value in order to ensure coincidence or convergence between the actual measured weight value and the target weight value. CONSTITUTION:For the steps of gate control using a computer, the weight value Wn=Wn-1+DELTAWn-1 is first obtained as shown in a figure (a) by synthesizing an error Wn-1, i.e., the difference between the target weight value W0 and the actual measured weight value W'n-1 obtained in the preceding measurement and the cutoff weight value Wn-1 obtained at the time point of the cutoff command in the previous measurement. Then a gate closing signal is delivered as shown in a figure (b) when the detected weight value is coincident with the above- mentioned synthetic weight value Wn in the measurement of this time. As shown by figures (b) and (c), an error Wn, i.e., the difference between the value W0 and the actual measured value W'n is obtained in the measurement of this time to decide whether the next measurement is started. When the next measurement is started, the step of the figure (a) is reset.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention is based on the calculation of 1iWo-W'n=Δ61 between the 1111 and the 0 accurate weight value WO generated in the lid scale and the true needle-side weight value lll'n, and the use of a calculator. According to the method rc of another car (ideally ll1O). The self-contained scale to which the present invention is directed should be widely understood in terms of its mechanism and the type of object to be weighed.

Here, we will discuss the errors found in general patch-type backing scales. In the figure of the cram school 1, (lmu) is the intoxication mechanism (]1 child detection sensor), ttS is the chute, (kiln 4 is the dart, 1141 is the gate barrier actuator, 0@ is the IIL bag, and n is the control mechanism.

Sensor (game at the moment when the river detects the amount of electricity wt) (Suppose that it sends II shut command No. 1!) There are the following four errors at this time.

(1) M Chishu mwVc includes the Ichimaru Iris amount f1 based on the military pressure of Kancho Rin 〇 (ll) I#jk Chiho-W VC is the chute exit and one bag in the sensing question. 1M4 difference corresponding to the needle tombstone floating in the 9th position with 411LtiIIt material decline l, ik
m f z is not included.

(1) According to the sensing law, the sensing value 11WVc includes the I#4 position kmfm corresponding to the to-be-received noodles that came out from the chute exit within the response delay time up to the moment when the gate closed IJNq. Not yet.

(JV) Sensing Jk lid W, from everyone who started the gate closing to everyone who closed the gate 2) When I left - several people who responded to the tumor that came out of the exit. irises f
My turn was not told.

Thus, when the gate completes its closing movement as shown in the g-th diagram, the ridge 1 (accommodated 7c4i-↑tumor chammen W' is expressed by the formula, W' = W ft + fs + fs + fa)
■It is given by. That is, -white tA minister E (F-)t
yo, formula, K = W'-'N = -110ft+is10ts
is given by Usually fs<h+fs+fa.

Rimari g=w'-w>o. For example, Kajugai W
′ is set to match the second wax (set value) No., then the sensing quantity (
l! side value) W is W' = Wo (W means you have to cut quickly. To put it simply, if you want to measure 80Kf, for example, design it so that it outputs a shut signal at B9.8. Figure 8) Figure 2 shows the total difference of 1 E.

#vJIL lid value (sensing irises) W and true needle IM in the cram school 4 diagram
The relationship between the ljm cover value (actual weight I1) W', the Ippon Arashi Tomb 11 (mokuchirowa) We, and the WA difference ΔW is shown.

Next, the conventional w4 difference correction force method will be explained.

Adjust the scale to Wo on the setting device 0 We on the scale is actually W as the mechanism of the scale. @6
See diagram. This setting is based on a rule of thumb, and with this setting, if the NIgJ command is issued at time t of W, it is expected that W o V will be exactly the same after retracing the progress of 8 B).However, there are various factors. The reality is that the process as shown in (b) results in W'. When - occurs in # like this, conventionally, ΔWt- is calculated, and in a basket of 100% or 80- of this, Reset.
This correction is of course done with help (artificially). 1
Since correction at 0G- often results in overshoot (excessive correction), the energization is corrected at 60- or 60-. 60- It is written in the figure that the fussho at the time of the correction is called.

This is called gyobata sodome. It's based on experience, and more on the gut than on the desk.

However, it is true that it lacks credibility because it is based solely on human knowledge. In other words, even with modifications, it is still m2i! In most cases, this will occur. In addition, when the shape of the needle cover object changes, aII'i again occurs and correction is required, but this -kaede causes the rudder NIh low F to a large extent.

The purpose of the present invention is to solve this problem of lack of efficiency and decreased efficiency.

The above-mentioned threshold hearing point fI: Solved Next, the structure of the Heisei era in which the present invention was developed, that is, the eleventh aspect of the present invention, is as follows.

This is 151! to implement the method of the present invention! Let's explain the structure of the scale that should be compared with reference to Fig. 6 of the scale diagram and the theory of operation Ij'1 Fig. 7 of the scale.

@Phantom is chute, (2) is gate, g11 line gate opening/closing actuator, (goods) is hopper, (to) is 1dong centimeter, (to) is amplifier, same 1jAD converter, - is calculator, 4 is shutter, - is the shutter opening/closing actuator, @
0 is a distance III measuring sensor that uses ultrasonic waves, etc. 4 is an amplifier. Let the computer (c) control the next step OK) In the chopping IJ1 turtle Wit-, at the time of the chopping command in the 1st column of the IIJ time [refer to the 7th bin], the , Mejiao weight grudge Wo and ^'s drunkenness #1 Huang Tomb 111 W'n
-s, that is, the error JWn-t (=Wo-W'n-
h)t&aL7tlljl1mWn-Wn-x+jW
Find n-tt. Jukumu jt value W (when it matches I, gate close elJ output 1 signal is output. $87 figure P) or d#
The difference between the l:IIIA heavy grave value Wo and the true *-J jib- value W'n, that is, the one-sho ΔWn, in the iIt# of the mouth is made longer. -Wn is O, i.e. w'
n=w.

The case where this is the case is shown in Figure 7 (b), and when ΔWn is not 0, that is, In'(Wn7) - is the case.

)Next L111 tt #j VC* ffc Capture the car 〇冑 When it moves, urinate on the step (A).

In the next IWIi-, -7-), Wfl+t=W,
1ΔWs>t″#11#11 Rinom Sumire side, Kari Iv
4 Sho Δvtn Shi 1 will last.

In the present invention, the following items (1) to (2) must be compared.

ψ The 11 round cuts mentioned in step U) of the pedestal structure
direct Wa-1 and miJ 114I2) d4 difference ΔIn
-1JV 疋-0 Next issue (1) and (including illusion).

Width pot d diagram bifu revise. The box MIJ times for the second time, only the (K-1)th time υ, that is, the only next Wk-
1 and ΔWk-, when 6 Wo-, -=w
m-1. Δ1vn-□=Δ11-(1) Figure 8-
)reference. Multiple 1g1(
11 times or more) W k −1*W k −1+
......Wk-m and ΔW k −
1+JWk-H+・・・・・・・・・・・・・・・
....

When referring to the total of ΔWk-m ◎In this case, wn
-, and ΔIn-1 are generally taken as arithmetic averages.

In other words, for -In-1, ΔVl n -1= (W k -1+W k -m
+=”−・10W k −” )7m ■ However, one method is to use the geometric mean.

1s4w□-t K ′IA ”C14 eba・
↓Δ*n-1=(Wk-1XWk-, X-・-・-・・
XWk-m)m
)Ck-1)・・・・・・・・・・・・)Ck-m and Ck-80ck-*10・・・・・・・・・・・・+C
k-m=1)' is preferable◎ For example, ΔI
When calculating n-1 by an arithmetic average, ΔN n -1" (Ck -1・W k -r
k −1・W k −t 10・・・・・・・・・・・・
10ch-m-wh-m) 7m Furthermore, W r* -1fC
Therefore, Wk-1 for only the vLL12 (h-8) song may be used (11-).

= Wk-1) ■ Also, in this (phantom case), the next (K+1) time l
In tt#J, -m+1.・・・・・・・・・
・、、、、・、 K -11K11l! If the sum of l
It is preferable to calculate 41L pieces of data for the dowel m(2) even if it is m 1g1CmiJ times. Multiplication table 11g1
Since it is rare that there is no data for one time, it is sufficient to enter appropriate data in 7 depending on whether it is given or not. Also, when K1t-j is 1t-j, if the value of K is small, such as 2 or 8, calculate JIM on data 1 of all preceding times. It's a pull. Also, it is preferable to have a VC card so that you can enjoy the mv box. m used in the first set
Instead of adding K(2) fine data to the data of Ckfl) in the first scooping and deleting the (k-m)-th data, no matter how large k becomes, One method is to use all the data from the circle to the (k-1)th piece. However, this can be considered a special case.

The effects of the present invention compared to the conventional method and the D ratio of 12 can be summarized as follows: (1) One degree of ItiIi'M in each round can be converted into a large ring.

(ii) The effort required for artificial modification and ILLA can be whispered.◎Next, we will introduce the appearance 1 of Katsuo & I to the strength of the present invention.

11J　A■ Re-plot the formula.

w' = w - f1 fs 10 fi 1 Is Once the hard interval of the scale is determined, f4 is determined.

Kakugekatsu single pressure and @Gangtan Izuigai Q
Considering the relationship between Flow F Shosa h (6th cause transduction) @J
l! WJ children's improvement is equal to impulse.

flt--m(VB-Vl) Transforming this formula, '+= (V+ V@) Friendship mt is @-quality cover v1=〆I1gh: llrg4+i Velocity before V, kneeling 0: Velocity immediately after collision Here 1, = J Q + K, t = 〆ggk,
It is called KIk "gli fist arrest rate constant".

(2) About the accent jlh after the shutdown.

This is 1z-Q 1
．． It represents the time it takes to reach ? Knee tLfK, (-
Articles of Incorporation.

lz -gt Q From the above, it can be seen that in order to know the -difference, it is sufficient to know the instantaneous mass flow lid Q and the falling height bl. A sensor is installed as shown in Fig. 6 in order to know the hl.

However, even if this is useful as Keigo's data for the 1111th fight, it can be made useless for the S-th and subsequent battles.
In other words, the correction that has already been added to kk may be executed at the step top.

To round to know Qt, it is sufficient to divide the difference between result-weight wt-/sWt at one particular instant during the calculation by the time difference Δ(.

9 companies (lft-wmt)/Δt However, since this is also executed in the correction [step 44], such calculation is not essential.

However, it goes without saying that it is never invalid to obtain h:Ql every time or every moment for the 6th rule as reference data.

(31 Determined by the hardware configuration, s f, ``delayed articles of incorporation'' K for both.

Above, L"K*' Ksk is synthesized to form "Taneai Separate Articles of Incorporation"
Define K. As K = xl + K * 10 Ka and KQ - = ΔW, the error t, t set value w1 and Δ
wVr reduced weight w w w, -4wt sensed W4t
Just send out the flJK gate shut signal gt.

The above is theoretically possible. 3. An example of a Glock circuit based on such theoretical considerations.
- The flowchart thereof is shown in FIG. 119810 in FIG.

131F1 weight detection sensor (load cell, etc.) is a Wheatstone bridge.It is one of the Rei・P-U@14 central! & physical equipment (commonly known as C,?・υ) (#shi< is Moto Coola's central processing It is also good to use a general C.p-13 instead of M-P-U K (which is a device).

FIG. 11 shows the test data for the trunk, which was planned using the conventional method with Ja- as the target value and a two-way system. The object to be measured is rice. The temperature was 1°C and the humidity was 70°C. Midfielder
The formula when il"71 is Q--X l O(fy4
The solid line is the data for the left scale, and the dashed line is the data for the right scale.

Figure 12 Fi is test data obtained when the method of the present invention was carried out under the same conditions. Left arithmetic mean i -8a0006
．． Standard − Jlz=laaG, right arithmetic mean X Ja(L
OGII, standard-difference *-1485, 6 points. The data direction is considered to be due to an abnormal external IL, and this can be ignored.

In the conventional method, the linear average error was ±IIOF, but when the method of the present invention was overused, the average error was ±mop, indicating a significant improvement in p1 measurement accuracy.

[Brief explanation of the drawing]

Figures 1 through 1G explain the conventional method, and Figures 13-I.
WA company@schematic configuration diagram, Figure 8 is an error vector diagram, Figure 4 is a diagram, and Figure 5 is a model diagram. II @ Figure or 11
Figure 10 is related to the present invention; Figure 6 is a schematic diagram of Fim;
@'1 Figure U) or D) is a diagram showing the original design, Figure 8 If) =- is a block diagram, Figure 9 is a block circuit diagram,
Figure IIIO is a flowchart, Figure 11111, No. 18
The figure is a diagram of experimental data.

Claims (1)

[Claims] Consisting of the steps set forth in (g), which operate the gate by opening the gate and converge the true drunken violet into the eye reef WO. Automatic correction - automatic correction force method using the vibration of the scale needle. H) At the time of the shredding command in jli#J of @s, the thin storm noodles Aki Wn-, in, -ri's It koji', the till jmu-1 shift of ^ with m-aki wo and ^. -s and Sho1wara l1g[ΔWn -H(=Wo-"lIn-1
) 1/1 Grave Basket W n=* n −t+jWn
Find -+. (L') On the day Z) At the 11th measurement, the result was 1.
When the j record formation Jlk grave 1wLW n LIC- is opened, the gate is closed and the 1st crop is biased. ? Today's IIF $113 VC, Eye - Emperor -
1st birthday wo and ^'s 1 "to prolong the ㆍmu mlμWn and υ seat, that is, the vassal JWn) to capture whether 9 will move to the needle side next time. (e) to step when it is a moving friend) Choose.