JPH032629A - Counting scale - Google Patents

Abstract

PURPOSE:To confine errors within an allowance range and to improve counting efficiency by issuing warning when a number (k) obtained by counting samples exceeds a target number N and when a maximum counting error I included in the number (k) exceeds an allowable maximum error E. CONSTITUTION:The following values are inputted through a keyboard 5: a single weight value mu of a product to be counted; a sample number (n); the coefficient epsilon of variation of the products; a target number N in counting; and an allowable maximum error E for N. The number (n) and the coefficient epsilonare used, and the sequence of a countable maximum number Li is counted at a probability P or more within (i) number of counting errors. The result is stored in a RAM 23. A number (k) is computed from the data that are measured in a weight detecting part 1 and the single weight value mu. The maximum counting error I included in the number (k) at the probability P or more is obtained. When the number (k) exceeds N and the error I exceeds E, a LED 4 is lit and warning is issued. Therefore, the fact that the error is within the allowable error range is ensured even if the counting is performed without noticing the number.

Description

DETAILED DESCRIPTION OF THE INVENTION <Industrial Application Field> The present invention relates to a counting scale that determines the number of identical articles, such as electronic parts, mechanical parts, or office parts, from their weight.

<Prior Art> In a counting scale, generally, the unit weight value (average weight) of a known number of items is determined from the weight of that item, and the number of items is determined by dividing the weight of an unknown number of items by that unit weight value.

By the way, according to such a counting method, a counting error occurs depending on the weight variation of the articles, and the magnitude of the error is also related to the number of articles counted at one time.

Therefore, when weighing a certain target number, the items are placed on the weight detection unit several times until the displayed value reaches the target number. The final counting error will also vary depending on the number.

In conventional counting scales, the occurrence of such errors is ignored, or the maximum countable number L0 that does not cause counting errors with a certain probability or more is determined in advance. Some have taken measures such as loading the following numbers at a time.

<Problem to be solved by the invention> Conventional counting scales that ignore errors naturally have unreliable counting results, and cannot be used to load items less than the maximum number L0 that can be counted without error. , the efficiency of counting work is poor.

In the task of measuring a target number of items, the maximum allowable counting error (or average allowable error) is often given for the number of items to be weighed. To realize a counting scale capable of efficiently performing counting work while guaranteeing that the number is within the range.

Means for Solving the Problem> In the present invention, as shown in the basic conceptual diagram of FIG. In a scale equipped with a piece counting means for calculating the number of pieces, the number of samples n used to calculate the unit weight value μ
and setting means d for the coefficient of variation ε of the article to be counted.
, input means C for manually inputting the target number N to be counted and the maximum allowable counting error E for the target number N, the coefficient of variation ε, the number n of samples, and the number calculation result k by the number calculation means. Counting error calculation means for calculating the maximum counting error ■ included in the number k with a given probability or more; and a comparison means g for comparing n and I and E, respectively, based on the comparison results. If k > N or I > E, . ), the next target number is set to N-k, and the next allowable maximum counting error is set to El, so that counting can be continued.

In this specification, the sample number setting means C and the variation coefficient setting means d are means for inputting the number of samples n or the coefficient of variation ε from the outside, or the scale itself calculates the unit weight value. It may also be a function of determining and storing the number of samples or the coefficient of variation ε of the articles to be counted during the process.

Function and principle of the invention> The number of samples n used when calculating the unit weight value μ of the article,
If the coefficient of variation ε of the article is known, the maximum counting error 1 included in the number k counted every - degree can be calculated with a probability P given in advance or higher.

In other words, from n, ε, k, the probability p that the counting error is O
. , similarly calculate the probability P that there will be j pieces (j is any integer), and p. Or Pa + S (p j + p - =) d (where I is a positive N number) I does not become smaller than the pre-given probability P
When , I becomes the maximum counting error included in k.

In other words, If Po≧P then I-0 and if not, 11) o+Σ(P;+p-J)≧P #II+ Let I be the minimum i that satisfies the following.

In this way, the maximum error (■) that is probabilistically expected to occur for each -degree count is determined, and this 1, the previously input allowable maximum counting error E, the target number N, and the current counting result. k respectively, and if (2) is satisfied, set the next target number and allowable maximum counting error as N-
Similarly, the following counting operation is possible as E-r and E-r,
In addition, in the case of (1), by notifying the operator of overloaded items, it is possible to ensure that the final counting error is within E, and to ensure that the final counting error is within E. The number of samples to be counted can be increased as much as possible within the allowed counting error.

<Example> FIG. 2 is a block diagram showing the configuration of an embodiment of the present invention.

The weight detection section 1 includes a load sensor that engages with the plate 1a and A-
It has a built-in D converter, etc., and can output digital data corresponding to the load on the plate 1a.

The data from the weight detection section 1 is stored in the CPU 21 and the ROM 22.
It is incorporated into the microcomputer 2 equipped with a RAM 23 and the like.

This microcomputer 2 includes a display 3 for digitally displaying the counting results, a light emitting diode 4 for notifying the operator of overloading of articles on the plate 1a, and a target number N and a tolerance. A keyboard mainly consisting of a numeric keypad for inputting the maximum counting error E, the unit weight μ of the articles to be counted, the number of samples n used for calculating the unit weight μ, and the coefficient of variation ε of the articles to be counted. 5 is connected, and an input/output port for connection with external equipment such as a printer or a personal computer is also provided.

The ROM 22 of this microcomputer 2 has programs written in it as described later, and the RAM
23 has a work area and an area for storing the above-mentioned various manual data from the keyboard 5, as well as a maximum (1? (Ll (i=O,l,
2,3. ...,) is set.

FIG. 3 is a flowchart I- showing the contents of the program written in the ROM 22. Hereinafter, with reference to this diagram, the operation of the embodiment of the present invention will be described along with the method of use.

Prior to the counting operation, the unit weight value μ of the articles to be counted, the number of samples n used to calculate the unit weight value μ, the coefficient of variation ε of the articles to be counted, the target number N to be counted, and the target number N The allowable maximum counting error E for the input value is entered manually using the keyboard 5.

Note that this human power may be used to transfer data from, for example, a personal computer connected to the input/output capotro.

Now, when such various data are manually generated, using the number of samples n and the coefficient of variation ε (A is sufficiently close to 1 given in advance b), the probability is greater than or equal to P and the counting error is within i pieces. The maximum countable number is the sequence [L;l (i=
o, 1.2, 3. ) and convert the sequence to R
Store in AM23.

The method for finding the sequence (L,) is as follows.

From the normal distribution table, β becomes. seek.

At this time, the maximum number that can be counted with probability P or more and within counting error i is determined as the largest integer that satisfies the following.

becomes. Note that () is a Gauss symbol and means the largest integer that does not exceed the internal value.

This sequence (L,) is 0<Lo<L, <LZ<,,,,<Li<,,,.

However, for example, if the result of counting an arbitrary number of samples is k, then L, -, <k≦L. ...(6)
10 is uniquely determined, and in this case, the counting error is 1
The probability that the number is 0 or less can be said to be greater than or equal to P (this does not mean that the counting error is 10).

Now, when the object to be counted is placed on the plate 1a after obtaining such a number sequence (shi,), the increasing type on the plate 1a is determined based on the data from the weight detection section 1, as in a normal counting scale. MW is calculated, and then the number k of articles is calculated by dividing the weight 1w by the unit weight value μ in the RAM 23 and rounding.

Then, the number k is compared with the target number N, and if k exceeds N, the light emitting diode 4 is turned on to notify that fact. If k is less than or equal to N, proceed to the next step, and from that number and the sequence (L,), find the number k with probability P or more.
Find the maximum counting error I included in (corresponding to 1° in equation (6) above).

Next, compare this I and the allowable maximum counting error E, and if 1 is E
If the load exceeds the limit, the light emitting diode 4 is turned on to notify the operator that the load is too high.

If I is less than or equal to E, the target number is set to N-k, the maximum allowable counting error is set to E-1, and the next addition of the items is waited.

According to this embodiment of the present invention, the worker freely places any number of items on the plate 1a from the beginning of counting, and only when the light emitting diode 4 lights up does a certain amount of the item be judged as being overloaded. If the counting operation is stopped when the displayed value falls below the target value N, it is guaranteed that the error included in the obtained counting result is within the allowable maximum counting error E. .

In other words, if l exceeds E by adding items several times, it means that the sum of all 1's up to that point exceeds the initially allowed maximum counting error E.
In this case, there is a possibility that the error included in the current count display value exceeds the maximum allowable count error E, and the light emitting diode 4 is turned on only in this case. It is. Therefore, the light emitting diode 4
Unless it lights up, it is guaranteed that the final counting result will contain only an error within the allowable maximum counting error E, regardless of the number of pieces placed on the plate 1a at a time.

Therefore, the operator can place as many samples as he or she likes without worrying about how many samples must be placed.
When the light emitting diode 4 is lit, it is only necessary to unload the sample until the light goes out, and it is sufficient to continue the work in this way until the displayed value reaches N.

In addition to the above-described lighting of lamps such as light emitting diodes, the means for notifying overloading may include sounding a buzzer or the like, or flashing a count display value.

In addition, the number displayed on the display 3 is the total number,
For example, to notify of overloading, "Sub, x"
It is also possible to switch the display to "0ver x" or the like. Furthermore, such an indicator can also be provided in conjunction with the total indicator.

By the way, by analyzing the above-mentioned a priori information of the maximum number of items that can be counted within counting error i in more detail, we find that the probability of occurrence of a counting error of 0 is q.

+1 piece q+ 1 piece 〃 q +1 (solid 〃 q・1 piece 〃
Find q (usually Q-i・Qi due to symmetry), and then e
+=O・Qo+1・(QuQ−+)L・・+t・(q
+tQ-υ・2ΣJQj is calculated as the average counting error e instead of the maximum counting error i (. Then, if you perform the same operation as in the previous example with E −e = 6 instead of It gets even better.

Alternatively, since the counting error can become larger depending on the efficiency, all three can be mixed, e. It is possible to adopt a method in which a limit is set in advance on the number of times that is used, for example up to three times, and ten is used from the fourth time onwards.

Further improving this, e, is applied only when the value of e exceeds a certain value (for example, 0.5). You may limit only the number of times that .

In addition, instead of formula (6), the usual expected value 0・qo+l・q++(−1)・q−+10・・−1−q
If you calculate t+(-4)・q,・(7), it becomes 0, so it is useless.

Furthermore, in each of the above embodiments, for safety, the target number N may be corrected to N10B, and N+E may be counted.

Furthermore, even if the maximum allowable counting error E is not input, the above-described operation may be performed automatically by setting a fixed number, for example, E=5.

When determining this E, do not always make it the same number, but depending on the target number N, N≦1000...E=5 1000<N≦toooo...E-10, etc. It may also be determined based on certain rules in relation to.

In addition, the number of samples n used for calculating the unit weight value should be set especially if the scale is of a type that automatically calculates by placing a predetermined number of samples on the counting scale. There's no need. Also,
Regarding the coefficient of variation ε, there is no need to input it from outside if the scale has a function to automatically calculate it by placing pieces one by one on the counting scale in the process of calculating the unit weight value, for example. .

<Effects of the Invention> As explained above, according to the present invention, even if the counting operation is performed without paying too much attention to the number of pieces to be loaded at one time until the overloading is notified, the final counting result will not change. is guaranteed to include only errors within the range of counting errors specified to be allowed.

Conversely, the operator can set the maximum allowable counting error as 1, 2, etc., and get an idea of what the allowable counting error is that he/she is most comfortable with.

By doing so, the operator can decide whether it is better to count more carefully or whether this base is sufficient.

[Brief explanation of the drawing]

Figure 1 is a basic conceptual diagram showing the configuration of the present invention, Figure 2 is a block diagram showing the configuration of an embodiment of the present invention, and Figure 3 is the ROM.
22 is a flowchart showing the contents of the program written in the computer. 1... Weight detection unit 2... Microcomputer 21... CPU 22..., ROM 23... RAM 3... Display unit 4... Light emitting diode 5...・Keyboard Fig. 1 Fig. 22

Claims (1)

[Scope of Claims] A scale equipped with a number calculation means for calculating the number of articles by dividing the weight value measured by the weight detection section by the unit weight value of the article, which is used to calculate the unit weight value. Number of samples n
a means for setting a coefficient of variation ε of the article to be counted;
Using an input means for inputting the target number N to be counted and the allowable maximum counting error E for the target number N, the coefficient of variation ε, the number n of samples, and the number calculation result k by the number calculation means, counting error calculation means for calculating the maximum counting error I included in the number k with a probability greater than or equal to the probability that If the above comparison result is k≦N) and I≦E, the next target number is set to N-k, and the next target number is set to N-k. A counting scale characterized in that it is configured to be able to continue counting with the maximum allowable counting error as E-I.