JPH0132927B2 - - Google Patents

Description

[Detailed description of the invention]

The present invention relates to counting scales. Errors that occur when counting multiple samples using a counting scale include errors in the scale itself and variations in the individual weights of the samples to be counted. In order to minimize the error caused by the latter, the conventional method is to correct the average weight, which should be the standard, according to the total number of samples each time an additional sample is counted (Japanese Patent Application Laid-Open No. 141072-1982). Or, when correcting this average weight, compare the temporary average weight and the average weight,
Determine whether the two match within a predetermined accuracy,
A counting scale (Japanese Unexamined Patent Publication No. 70421/1983) that issues a warning signal if they do not match is used. However, with this method, if the number of samples added during additional weighing and counting is too large, the number of pieces obtained by dividing the weighed value by the temporary average value may contain an error close to an integer. may contain one or more errors. In such a case, it is meaningless to perform any kind of "rounding" operation on fractions less than 1 of the counted number, or to examine the consistency of the difference between the tentative average weight and the average weight. No errors can be detected. Therefore, even if the average weight is corrected based on the value obtained in this way, counting errors cannot be prevented. The present invention aims to eliminate the above-mentioned deficiencies in the counting method and to provide a counting scale that controls the occurrence of errors so that one or more errors do not occur during each count. A configuration for achieving this purpose will be explained with reference to the basic conceptual diagram shown in FIG. The present invention includes the following (a) to (l), and only when the calculated additional number a _j is equal to or less than the allowable maximum additional number N' based on the determination result by the determining means (l). It is characterized by being configured so that the number a _j is added to the cumulative number. (a) Electronic scale section (b) Divide the weight W ₁ obtained when a known number n ₁ of samples is loaded into the electronic scale section (a) by the number n ₁ to obtain the first temporary average weight m ₁ . First provisional average weight calculating means (c) Calculating the increased weight W _i (i = 2, 3...s) when adding an unknown number of samples to the electronic balance section (a) Temporary average weight m _i-1 of (i-1)
(d) A first number calculating means for calculating the additional number n _i by dividing each increased weight W _i by (i = 2, 3...s) _.
The i-th provisional average weight m _i (i=2, 3...
s) Second provisional average weight calculation means (e) Adding means for calculating the cumulative number by sequentially adding the above additional number n _i to the known number n ₁ (f) Displaying the cumulative number by the adding means (e) Display (G) Total weight _ΣW (=W ₁ +W ₂ +
...+W _s ) by adding means (E) to the cumulative number N (=n ₁
Average weight calculation means ( _H ) that calculates the average weight by dividing by _s temporary average weights m ₁ and m _i and the average weight _. Variation coefficient calculation means (l) Permissible maximum additional number calculation means (n) that calculates the allowable maximum additional number N' whose counting error falls within a predetermined value of less than 1 from the above cumulative number N and variation coefficient C _v . A second number calculation means (L) that calculates the additional number a _{j by dividing the increased weight W j} when an unknown number of samples is added to the electronic scale section ( _A ) by the average weight after calculating the number N'. Calculated number of additional pieces a _j and maximum allowable number of additional pieces
Judgment means for judging the magnitude relationship of N' Examples of the present invention will be described below. FIG. 2 is a block diagram showing the configuration of an embodiment of the present invention, and FIG. 3 is a flowchart showing the contents of the program written in the ROM 4. In FIG. Note that the Roman numerals (,...) are numbers indicating predetermined stages of the flowchart. In FIG. 1, the load placed on the tray 3 of the electronic scale section is converted into a digital electrical signal and output. The CPU has a built-in arithmetic unit, clock signal generator, etc., and controls each peripheral device via a bus line. ROM4 stores programs,
The RAM 5 temporarily stores measured values, calculated values, etc. These CPU, ROM, RAM, and bus line constitute the microcomputer 11. Further, the first display 1 displays the cumulative number of samples to be counted, and the second display 2 displays the number of samples to be additionally weighed. The keyboard 12 consists of a reference memory key 6, an average update key 7, a tare key 8, an average clear key 9 and a numeric key 10. First, when a predetermined number n of ₁ pieces, for example 5 pieces, is placed on the tray 3 of the electronic scale and the reference memory key 6 is pressed, its weight W ₁ is divided by 5 according to the program stored in the ROM 4. , first provisional average weight
m ₁ is stored in RAM5. The first display 1 indicates the number of additional samples, 5, and the second display 2 indicates the number of additional samples that approximates 5 (). Add an unknown number of samples (4 to 6) close to the specified number to the electronic balance pan 3 () and press the reference memory key 6, or
Alternatively, by automatically storing after automatic stability detection, the increased weight W ₂ is divided by the temporary average weight m ₁ and the additional number rounded off to the nearest whole number is n ₂ , for example, 5, and the above The second provisional average weight m ₂ obtained by dividing the increased weight W ₂ by this counted number n ₂ (=5) is
Stored in RAM5 (). If the number of additional samples is 2 or more or less than the specified number of additional samples (in this case, 3 or less or 7 or more), it is desirable to issue a warning and prevent m ₂ from being calculated and stored. This is because in equation (1) shown later, n is a constant value (for example, 5), and if this value deviates significantly, the error will increase. However, since limiting n completely to a constant value would place a heavy burden on the workers, it is appropriate to allow a range of ±1. The first display 1 displays the sum n _{1 +} n 2 (=10) of the previously displayed number n ₁ (=5) and _{the current counted number n 2 (} =5) (), Display 2 at 2 again indicates the additional number of approximations to 5. Unknown number close to this specified number (4~
6) samples were added to the electronic balance dish 3 again,
By repeating the above process a predetermined number of times, for example, 5 times (), a total of 6 temporary average weights m ₁ ,
m ₂ , . . . m ₆ are stored in the RAM 5, and the first display 1 displays the total number of samples 30 thus far. The second display 2 has the following formula However,; Total weight / Total number of pieces N = Total number of pieces (=n ₁ + n ₂ +...+n ₆ = 30) n = Initial number of pieces weighed (= 5) Ds = Standard deviation of provisional average weight (=√ variance) Derived from , and displays N′ in equation (1) below. In addition, to explain the following formula, the first left side
The term is an error amount due to an error on the reference weight side, and the second term is an error amount due to an error on the counted sample weight side, and the entire left side is the statistical addition value of these. If this is 1/2 or less of the average weight, no error will occur in the rounded count value, and the right-hand side is a term that indicates this condition. However, the coefficient 3 before Ds is calculated with a probability of 99.7%. Here, since Ds is the standard deviation of n provisional average weights, there is a relationship between Ds=√·SD and the standard deviation SD obtained when measuring the weight of each sample. The first and second terms on the left side are
If expressed using SD, each

[expression] and

[Formula] becomes. And what these mean is as follows. Now, let SD be the standard deviation of the variation in sample weight,
When the average weight of N samples is calculated, the deviation of the calculated average weight from the true average weight is with a probability of 99.7%.
(hence 3SD) under It is expressed as If we count N′ based on this, we get
When counting N' samples, the following error occurs in weight units: This corresponds to the first term on the left side. On the other hand, when N' samples are placed on a plate, their average weight is relative to the true average weight. It's off by just that. Therefore, the error in the total weight is This corresponds to the second term on the left side. Now, by modifying the above formula, we can calculate the upper limit of the number of additional pieces to make the counting error less than 1, that is, the maximum allowable number N'=(1/6Cv) ² 1/n N/N+N'...(1) However, Cv= The coefficient of variation = Ds/W is displayed (). Further, the average weight obtained by dividing the total weight by the total number of pieces is stored in the RAM 5. If the coefficient of variation Cv is determined to be, for example, 0.0045 in the above, the value of N' according to equation (1) is 76.9, and the value 76, which is obtained by rounding down fractions less than 1, is displayed on the second display 2. be(). 76 according to this display
If the number of items within the range, for example, 70 items, is added to the pan 3 of the electronic scale, the increased weight will be divided by the average weight by pressing the average update key 7 or by being automatically updated after automatic stability detection. The additional number of pieces is rounded off to the nearest 1 and calculated as 70 (), which is added to the 30 pieces above, making the total number of pieces 100.
is displayed on the first display 1 as (). In addition,
The error estimated for the number 70 counted above is ±1/2
It is reasonable to consider that the above-mentioned total number of 100 so far is a value counted with zero error until it reaches 99.7% according to equation (1). Next, the calculation using equation (1) is performed again (), N′=(1/6×0.0045) ²・1/5・(100/100+
The number 123 obtained by rounding down the fraction of N') = 123.01 is displayed on the second display 2. This number 123 is the maximum allowable additional number calculated based on the current cumulative number of samples on the dish 3, 100, which is considered to have zero error, that is, the maximum allowable enlarged additional number. Here, by pressing the average update key 7 or by automatically updating after automatic stability detection, the second average weight ₂ is obtained by dividing the total weight by the total number of pieces, and the previous average weight (= ₁ ) is corrected. This value ₂ was determined using the total number that is considered to contain no errors, and is closer to the true value. The above operation process is repeated until the desired count is reached. In addition, in the final addition stage, if one counting error is acceptable, it is possible to add up to four times the number indicated by the second display 2. The reason is that when one error is allowed, 6Cv in equation (1) changes to 3Cv, and (1/3Cv) ² =4(1/6Cv) ² . Also, if 6Cv in equation (1) is changed to 9Cv, the number of pieces will be judged with an error of 1/3, and if it is set as 15Cv, the number will be judged with an error of 1/5, improving accuracy. N' decreases, and the number of weighing operations required to count the required number increases. The counting accuracy can be selected as necessary, taking into consideration the balance between the two. After counting the desired number of samples through the above operation process, when the sample is unloaded from the pan 3 of the electronic balance and counting of the same number is started again, the display on the first display 1 returns to 0, but the previous counting process Since the memory of the average weight obtained in step 1 is maintained as is, counting operation can be started immediately from the last added number. After that, divide the increased weight by the average weight,
The additional numbers obtained by rounding off the fractions are sequentially added and displayed on the first display 1. On the other hand, the second
The display 2 continues to display the first addition number (the final addition number in the previous counting process). If you want to count more pieces than last time, return to the measurement routine in ○A in the flowchart.
The average weight will be updated and the number that can be added will increase. This is because if you immediately return to the routine marked with ○ after this counting operation, the total number of items placed on the plate 3 of the electronic scale will be small, so updating the average will have the disadvantage that the number of items that can be added will actually decrease. be. As described above, according to the present invention, the average weight is sequentially corrected within the range of variation in sample weight, and the number of samples to be added is specified each time within a range that does not cause counting errors. is obtained. As a modified embodiment of the present invention, multiple times
In the part where the coefficient of variation of the sample weight is calculated from the variation in the tentative average weight due to an almost constant number of samples (see page 7), the maximum Cv of the sample to be counted is limited to 0.03 or less (in general, Cv rarely exceeds 0.03). )
Pre-allowed increment number, e.g. 5 initially
pieces, then 10 pieces, third time 15 pieces, fourth time 18 pieces, fifth time
It is possible to calculate the temporary average weight of each sample while increasing the number of samples sequentially, such as 21 pieces, 22 pieces at the sixth time, and the variation coefficient of the sample weight from the variation. In that case, the coefficient of variation calculation formula is as follows: P ₁ , P ₂ , ...Pi, ...Ps for each number, and the average weight for each.
Q ₁ , Q ₂ …Qi, Qs =〓 ^S / _i=1 Qi /〓 ^S / _i=1 Pi. Note that di=(Qi−)√, and the average weight W is updated every time the number increases. Next, use the Cv obtained from the above formula to calculate the additional allowable number N′, but by replacing formula (1) with formula (2).
Calculate N′. N'=(1/6Cv) ² N/N+N...(2) After you have finished calculating Cv, you can simply return to the same routine as after measuring Cv before the modification, or even after returning to the same routine. There are two ways to proceed with measurement while improving the accuracy of Cv by repeatedly determining Cv. In any case, after the Cv is determined, the maximum allowable additional number is calculated according to the determined Cv. The advantage of this modified example is that the number of specimens is increased even in the step of calculating Cv, so when the Cv has been calculated, there are already 91 specimens on the plate in the above example, and the deformed Since this is more than the 30 pieces used in the previous method, the number of pieces that can be added is also larger, and the required number of pieces can be measured quickly. This modification is included within the scope of the present claims.

[Brief explanation of drawings]

Figure 1 is a basic conceptual diagram showing the configuration of the present invention;
The figure is a block diagram showing the configuration of an embodiment of the present invention.
The figure is a flowchart showing the contents of the program written in the ROM 4. 1...First display, 2...Second display, 3...Weighing pan of electronic scale, 4...ROM, 5...RAM, 6
...Reference memory key, 7...Average update key, 8...Tare key, 9...Average clear key, 10...Numeric key, 11...Microcomputer, 12...Key
board.

Claims (1)

[Claims] 1. A first method of calculating a first provisional average weight m ₁ by dividing the weight W ₁ obtained when a known number n ₁ of samples is loaded on the electronic scale section by the number n ₁ . Temporary average weight calculation means and increased weight W _i (i=2, 3...s) when an unknown number of samples are added to the electronic scale section.
is divided by the (i-1)th provisional average weight m _i-1 (i=2, 3...s) calculated immediately before, and the number of additional pieces is calculated.
a first number calculating means for calculating n _i ; and calculating an i-th provisional average weight m _i (i=2, 3...s) by dividing each of the above-mentioned increased weights W _i by the corresponding additional number n _i ; a second provisional average weight calculation means for calculating the cumulative number of pieces, an adding means for sequentially adding the additional number n _i to the known number n ₁ to obtain a cumulative number, a display for displaying the cumulative number by the adding means; After calculating the temporary average weight m _s , the total weight ΣW loaded on the above electronic scale section
(=W ₁ +W ₂ +...+W _s ) divided by the cumulative number N (=n ₁ +n ₂ +...+n _s ) calculated by the addition means to calculate the average weight;
A coefficient of variation calculating means for calculating a coefficient of variation C _v of the sample weight from the variations in m ₁ and m _i and the above average weight, and a maximum allowable counting error within a predetermined value of less than 1 from the cumulative number N and the coefficient of variation C _v . Additional quantity
Maximum allowable additional number calculation means for calculating N′ and increased weight W _j when an unknown number of samples are added to the electronic scale after calculating the above-mentioned maximum allowable additional number N′
a second number calculating means for calculating the number of additional pieces a _j by dividing the number by the above-mentioned average weight, and a determining means for determining the magnitude relationship between the calculated number of additional pieces a _j and the above-mentioned maximum allowable number of additional pieces N'. and, based on the determination result, the electronic counting scale is configured to cause the adding means to add the number a _j to the cumulative number only when the additional number a _j is equal to or less than the N'. 2. The average weight calculation means calculates a corrected average weight by dividing the total weight loaded on the electronic scale section by the cumulative number after the number a _j is added to the cumulative number, and The means of calculating the number of
Thereafter, the corrected average weight is used to calculate the number of pieces, and the maximum allowable additional number calculating means calculates the maximum allowable enlarged number of additional pieces from the cumulative number to which the number a _j is added and the coefficient of variation, and the determining means calculates the maximum allowable additional number of pieces. 2. An electronic counting scale according to claim 1, characterized in that the electronic counting scale is configured to supply the electronic counting scale to a user. 3. The electronic counting scale according to claim 1 or 2, further comprising means for displaying the maximum allowable additional number or the expanded allowable maximum additional number.