JP2009115740A - Combination balance - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a proper combination balance capable of uniformly suppressing dispersion in weight values of products constituting optimal combination in one of a plurality of weight ranks, and forms an optimal combination of each weight rank of a uniform ratio in a balanced manner. <P>SOLUTION: Each commercial product ranks R1, R2, and R3 as a weight rank are composed of combinations of proper product ranks r1, r2, r3, r4, and r5. Here, each commercial product rank R1, R2, and R3 and each weight value range Δw1, Δw2, and Δw3 are substantially equivalent to one another. Meanwhile, the products belonging to r2 are separated to the rank R1 and the rank R2 rank and sharing these at a predetermined rate, and the products belonging to r4 rank are separated to the rank R2 and the rank R3 rank that shares theses at a predetermined rate. Accordingly, among each of the commercial product ranks R1, R2, and R3, the distribution rates p1, p2, and p3 of product belonging to each rank become substantially equivalent to one another. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a combination weigher, and in particular, selects an article having a substantially normal distribution using a weight value as a variable according to a plurality of weight ranks according to the weight value, and various combinations of weight values of articles belonging to the same weight rank. The combination weight is selected so that the combination weight value is equal to or closest to the target weight value.

  As this type of combination weigher, there is a conventional one disclosed in Patent Document 1, for example. According to this prior art, first, a large number of articles of the same type having different weight values are sorted by a plurality of weight ranks in a sorting mode. In the combination mode, the weight values of the articles belonging to the same weight rank are combined, and the optimum combination in which the total weight value of the combination matches or is closest to the target weight value is selected. Thereby, the weight value as a whole is constant, and it is said that a product in which the weight values of the individual articles are uniform can be obtained.

Japanese Patent No. 2678169

  By the way, in the above-mentioned prior art, the case where three weight ranks R1, R2 and R3 are set is disclosed as a specific example. And as goods, what is called primary products, such as agricultural products and marine products, are assumed. In addition, it can be considered that the articles | goods which are primary products are substantially normally distributed by making a weight value into a variable.

  Here, it is assumed that the weight ranks R1, R2, and R3 are set as shown in FIG. 7, for example. That is, assuming that the weight value of the article is wx, the average value thereof is wm, and the standard deviation is σ, w1 = wm−3 · σ, w2 = wm−σ, w3 = wm + σ and w4 = wm + 3 · Four boundary weight values of σ are set, and among these, the region of w1 ≦ wx <w2 is R1 rank, the region of w2 ≦ wx <w3 is R2 rank, and the region of w3 ≦ wx ≦ w4 is R3 rank It is assumed that In this case, the weight value width Δw1 (= w2-w1) of the R1 rank, the weight value width Δw2 (= w3-w2) of the R2 rank, and the weight value width Δw3 (= w4-w3) of the R3 rank are all 2 · σ. Therefore, in any of the weight ranks R1, R2 and R3, the variation in the weight value wx of the articles belonging to each of them is 2.0 · σ or less. Therefore, the variation in the weight value wx of the articles constituting the optimum combination is also 2 · σ or less.

  However, when the respective weight ranks R1, R2 and R3 are set in this way, that is, the respective boundary weight values w1, for partitioning the respective weight ranks R1, R2 and R3 with respect to the weight distribution of the article having a substantially normal distribution. When w2, w3 and w4 are set at equal intervals, as can be seen from FIG. 7, there is a difference in the distribution ratio (relative frequency) of articles belonging to each of the weight ranks R1, R2 and R3. For example, assuming that the distribution ratio of the entire article is 1, the distribution ratio p1 of articles belonging to the end R1 rank is expressed by the following Expression 1.

<< Formula 1 >>
p1 = p [w1] −p [w2] = 0.4987−0.3413 = 0.157

  In Equation 1, p [w1] is a distribution ratio of articles belonging to a region between the boundary weight value w1 and the average weight value wm in FIG. 7, and the distribution ratio p [w1] is “0. The value 4987 ″ is derived from the standard normal distribution table of FIG. That is, when the boundary weight value w1 is normalized and expressed by the absolute value z1, z1 = | {w1-wm} / σ | = 3. Then, when this absolute value z1 is replaced with the variable z in the standard normal distribution table of FIG. 8, a value of “0.4987” is derived as the distribution ratio (cumulative probability value) corresponding to this. In Equation 1, p [w2] is a distribution ratio of articles belonging to a region between the boundary weight value w2 and the average weight value wm in FIG. 7, and “0.3413” of this distribution ratio p [w2]. The value "" is similarly derived from the standard normal distribution table of FIG. That is, when the boundary weight value w2 is normalized and expressed by the absolute value z2, z2 = | {w2-wm} / σ | = 1. Then, when the absolute value z2 is replaced with the variable z in the standard normal distribution table of FIG. 8, a value of “0.3413” is derived as the corresponding distribution ratio.

  On the other hand, the distribution ratio p2 of articles belonging to the central R2 rank is expressed by the following formula 2.

<< Formula 2 >>
p2 = p [w2] + p [w3] = 0.41313 + 0.3413 = 0.6826

  In Equation 2, p [w3] is a distribution ratio of articles belonging to the region between the average weight value wm and the boundary weight value w3 in FIG. 7, and the distribution ratio p [w3] is “0. The value 3413 ″ is also derived from the standard normal distribution table of FIG. 8 as described above. However, the value of this distribution ratio p [w3] is the same value of “0.3413” as the distribution ratio p [w2] described above, and the corresponding areas are symmetrical with respect to the average weight value wm. I understand that.

  The distribution ratio p3 of the articles belonging to the remaining R3 rank is “0.1574” equivalent to the distribution ratio p1 of the articles belonging to the R1 rank because the R3 rank is symmetrical to the R1 rank with respect to the average weight value wm. "

  In short, the optimum combination constituted by the articles belonging to the R1 rank and the optimum combination constituted by the articles belonging to the R3 rank are respectively generated at a ratio of 15.74 [%], whereas It is statistically expected that the optimal combination constituted by the articles belonging to the R2 rank is generated at a ratio of 68.26 [%]. That is, there arises an inconvenience that the optimum combination of the weight ranks R1, R2, and R3 cannot be generated in a balanced manner.

  In order to eliminate this inconvenience, it is assumed that the boundary weight values w1, w2, w3, and w4 are set, for example, as shown in FIG. That is, w1 = wm−3 · σ, w2 = wm−0.43 · σ, w3 = wm + 0.43 · σ, and w4 = wm + 3 · σ are set. In this case, the distribution ratio p1 of articles belonging to the R1 rank is expressed by the following expression 3 in the same manner as the above expression 1.

<< Formula 3 >>
p1 = p [w1] −p [w2] = 0.4987−0.1664 = 0.323

  The distribution ratio p2 of articles belonging to R2 is expressed by the following equation 4 in the same manner as the above equation 2.

<< Formula 4 >>
p2 = p [w2] + p [w3] = 0.1664 + 0.1664 = 0.3328

  Further, the distribution ratio p3 of the articles belonging to the R3 rank is “0.3323” which is the same as the distribution ratio p1 of the articles belonging to the R1 rank.

  That is, the optimal combination configured by the articles belonging to the R1 rank and the optimal combination configured by the articles belonging to the R3 rank are each generated at a ratio of 33.23 [%] and configured by the articles belonging to the R2 rank. The optimal combination is also generated at approximately the same ratio of 33.28 [%]. That is, the optimum combination of the respective weight ranks R1, R2, and R3 is generated with a good balance.

  However, in this case, naturally, the respective weight value widths Δw1, Δw2, and Δw3 are unbalanced between the respective weight ranks R1, R2, and R3. Specifically, the weight value width Δw1 of the R1 rank and the weight value width Δw3 of the R3 rank are 2.57 · σ, respectively, whereas the weight value width Δw2 of the R2 rank is Δw2 = 0.86 · σ. It becomes. Therefore, for the optimal combination of the R2 rank, the variation in the weight value wx of the articles constituting the R2 rank is 0.86 · σ or less, whereas the optimal combination of the R1 rank and the optimal combination of the R3 rank , The variation in the weight value wx of the articles constituting each becomes 2.57 · σ at the maximum, and the quality as a product is remarkably inferior.

  Therefore, the present invention can uniformly suppress the variation in the weight values of the articles constituting the optimum combination in any weight rank, and can generate the optimum combination of each weight rank in a balanced manner with a uniform ratio. It is an object to provide a combination weigher.

  It is also an object of the present invention to provide a sorting device premised on being used in such a combination weigher.

  In order to achieve this object, the combination weigher according to the first aspect of the present invention is the same as the sorting means for sorting the articles having a substantially normal distribution with the weight value as a variable, according to the weight value, according to a plurality of weight ranks. Combination selection means for selecting various combinations of weight values of articles belonging to the weight rank and selecting a combination having a combination weight value equal to or closest to the target weight value. The weight value widths of the plurality of weight ranks are substantially equal to each other, and each of the plurality of weight ranks has an overlapping area overlapping with other weight ranks.

  In other words, according to the combination weigher of the first invention, the articles as the objects to be weighed are sorted by the plurality of weight ranks by the sorting means according to the weight value. Then, the combination selection means variously combines the weight values of the articles belonging to the same weight rank, and selects the combination in which the combination weight value is equal to or closest to the target weight value, that is, the optimum combination. Here, the weight value widths of the respective weight ranks are substantially equal to each other. Accordingly, the variation in the weight values of the articles belonging to each weight rank is substantially uniform, and therefore the variation in the weight values of the articles constituting the optimum combination is also substantially uniform. In addition, each weight rank has an overlapping region that overlaps with another weight rank. Therefore, the articles belonging to the overlapping area are distributed to the respective weight ranks including the overlapping area. Thereby, the distribution ratio of articles belonging to each weight rank is made uniform.

  In the first invention, each weight rank may be composed of a plurality of small regions partitioned by a plurality of boundary weight values. In this case, some of the small areas are shared by two or more weight ranks as overlapping areas.

  Further, it is desirable that the sorting means sorts articles belonging to the overlapping area into respective weight ranks including the overlapping area at a predetermined sorting ratio.

  The selection ratio referred to here is preferably determined so that the distribution ratio of articles belonging to each weight rank is substantially equal.

  The first invention can be applied to, for example, a combination weigher provided with a plurality of weighing means for measuring the weight value of the supplied article while each article is supplied. In this case, in a state where the articles are supplied to the respective weighing means, the selection by the selection means and the selection by the combination selection means are continuously executed based on the weight measurement value by the respective weighing means. It is good.

  Moreover, this 1st invention is applicable also to the combination scale provided with the selection mode and the combination mode. In this case, first, sorting by the sorting unit is executed in the sorting mode. Thereafter, the selection by the combination selection unit may be executed for each group of articles belonging to the same weight rank in the combination mode.

  The second invention of the present invention is a sorting device that sorts articles having a substantially normal distribution with a weight value as a variable according to a plurality of weight ranks according to the weight value. Various weight values of articles belonging to the same weight rank are selected. It is assumed that the combination weight value is used in a combination weigher that selects the combination whose combination weight value is equal to or closest to the target weight value. Under this premise, the weight value widths of the plurality of weight ranks are substantially equal to each other, and each of the plurality of weight ranks has an overlapping region overlapping with other weight ranks.

  That is, the sorting apparatus of the second invention is similar to the sorting means in the first invention, and is used in a combination scale similar to the combination selecting means in the first invention. Under this assumption, each weight rank Is defined in the same manner as in the first invention. Therefore, according to the second invention, the same operation as the first invention is achieved.

  As described above, according to the present invention, the variation in the weight value of the articles belonging to each weight rank is substantially uniform, and the distribution ratio of the articles belonging to each is made uniform. Therefore, it is possible to uniformly suppress the variation in the weight values of the articles constituting the optimal combination in any weight rank, and to generate the optimal combinations of the respective weight ranks at a uniform ratio in a balanced manner.

  One embodiment of the present invention will be described below.

  The combination weigher 10 according to the present embodiment is called a manual type. As shown in FIG. 1, as shown in FIG. 1, one control device 30 and N () connected to the control device 30 via a communication line 50 via a bus. N: an integer greater than or equal to 2) measuring instruments 70, 70,... In addition, the value of the number N of the measuring instruments 70 is suitably about N = 10-20.

Among these, the control device 30 includes an operation key 32 as command input means and a liquid crystal display 34 as display means. More specifically, as shown in FIG.
Processing Unit) 36 is provided, and operation keys 32 and a display 34 are connected to the CPU 36 via an input / output interface circuit 38. In addition, by adopting a so-called touch screen as the display 34, the display 34 may have the same function as the operation key 32.

  Further, the communication line 50 is also connected to the CPU 36 via the input / output interface 38. The CPU 36 is also connected with a memory circuit 40 as a storage means. The memory circuit 40 is composed of a semiconductor memory such as a ROM (Read Only Memory), a RAM (Random Access Memory), an EEPROM (Electronically Erasable and Programmable Read Only Memory), and the memory circuit 40 includes an operation of the CPU 36. A control program (main program) for controlling the program is stored.

  On the other hand, each weighing instrument 70 includes a weighing platform 72 on which an article (not shown) as an object to be weighed is placed manually (manual operation by an operator). As shown in FIG. 3, the weighing table 72 is coupled to a load sensor 74. The load sensor 74 outputs an analog weighing signal having a voltage corresponding to the weight value wx of the article placed on the weighing table 72. Generate. For example, a strain gauge type load cell is suitable as the load sensor 74, but is not limited thereto.

  The analog weighing signal generated by the load sensor 74 is amplified by the amplification circuit 76 and then input to the A / D conversion circuit 78. The A / D conversion circuit 78 samples the input analog weighing signal at a predetermined sampling period and converts it into a digital weighing signal. The sampling period by the A / D conversion circuit 78 is suitably about 1 [ms], for example, but is not limited to this. Further, at the stage before the analog weighing signal is input to the A / D conversion circuit 78, for example, at the input side or the output side of the amplification circuit 76, a relatively high frequency band noise component included in the analog weighing signal, for example, 100 An appropriate analog filter circuit may be provided in order to remove noise components mainly due to electrical factors above [Hz].

  The digital weighing signal converted by the A / D conversion circuit 78 is input to the CPU 82 via the input / output interface circuit 80. The CPU 82 performs a moving average on the digital weighing signal in order to remove a noise component in a relatively low frequency band included in the input digital weighing signal, for example, a noise component mainly due to mechanical factors of 100 [Hz] or less. Appropriate digital filtering processing such as processing is performed. Then, the weight value wx of the article is calculated based on the digital weighing signal after processing.

  The weight value wx of the article calculated by the CPU 82 is transmitted to the control device 30 (CPU 36). For this reason, the CPU 82 is connected to the communication line 50 described above via the input / output interface circuit 80. The CPU 82 is also connected via an input / output interface circuit 80 with three types of lamps 84, 86 and 88, which will be described later, of “R1”, “R2” and “R3”, which have different emission colors. Further, a memory circuit 90 constituted by a semiconductor memory is also connected to the CPU 82, and a control program (subprogram) different from the above for controlling the operation of the CPU 82 is stored in the memory circuit 90. ing.

  Each weighing instrument 70, 70,... Is assigned an individual identification number n (n = 1 to N) for identifying each of the measuring instruments 70, 70,. Further, although not shown in the figure, the control device 30, the communication line 50, and the measuring instruments 70, 70,... Are integrated by being accommodated in one housing.

  According to the combination weigher 10 of this embodiment configured as described above, the combination weighing operation is performed in the following procedure.

  That is, first, an article is manually placed on each of all the weighing devices 70, 70,... (The weighing tables 72, 72,...). In addition, the articles | goods said here are comparatively large agricultural products, such as an apple and an onion, for example, and it is mounted one by one with respect to one measuring instrument 70.

  Each measuring instrument 70 (CPU 82) calculates the weight value wx of the article placed on itself in the manner described above. The calculated weight value wx is transmitted to the control device 30 via the communication line 50.

  The control device 30 (CPU 36) displays the N weight values wx transmitted from the measuring instruments 70, 70,... Together with the identification numbers n of the measuring instruments 70, 70,. At the same time, the control device 30 sorts each weight value wx by three product weight ranks (hereinafter referred to as product ranks) R1, R2, and R3. And the combination calculation for selecting the optimal combination for each product rank R1, R2, and R3 is performed.

  Specifically, for example, for the R1 rank, the weight value wx belonging to this rank is combined by a predetermined combination number M1. The combination weight value Wy1 is within the allowable weight range (Wt1 ≦ Wy1 ≦ Wt1 + ΔW1) with the lower limit of the target weight value Wt1 set for the R1 rank, and is the closest or equal to the target weight value Wt1. Are selected as the optimum combination. In addition, the width ΔW1 of the allowable weight range is about the number [%] of the target weight value Wt1.

  Similarly, for the R2 rank, the weight value wx belonging to this rank is combined by a predetermined combination number M2. The combination weight value Wy2 is within an allowable weight range (Wt2 ≦ Wy2 ≦ Wt2 + ΔW2) with the target weight value Wt2 set for the R2 rank as a lower limit, and is the closest or equal to the target weight value Wt2 Are selected as the optimum combination. The allowable weight range width ΔW2 for the R2 rank is also set to about the number [%] of the target weight value Wt2. Further, the number of combinations M2 in the R2 rank may or may not be the same as the number of combinations M1 in the R1 rank.

  Further, for the R3 rank, the weight value wx belonging to this rank is combined by a predetermined combination number M3. The combination weight value Wy3 is within an allowable weight range (Wt3 ≦ Wy3 ≦ Wt3 + ΔW3) with the target weight value Wt3 for the R3 rank as a lower limit, and a combination that is equal to or closest to the target weight value Wt3 is optimal. Select as a combination. The allowable weight range width ΔW3 for the R3 rank is also set to the number [%] of the target weight value Wt3. The number of combinations M3 in the R3 rank may or may not be the same as at least one of the number of combinations M1 in the R1 rank and the number of combinations M2 in the R2 rank.

  When the optimal combination is selected for the R1 rank, for example, by such a combination calculation, the control device 30 applies the above-described “R1” to each weighing instrument 70 on which the articles constituting the optimal combination are placed. For example, an instruction to turn on the yellow lamp 84 is sent. Then, the “R1” lamp 84 of the measuring instrument 70 receiving this instruction is turned on. When the “R1” lamp 84 is lit in this way, the operator selects the optimum combination for the R1 rank, and the article placed on the measuring instrument 70 whose “R1” lamp 84 is lit. Constitutes an optimal combination of R1 ranks. Then, the operator manually removes the articles constituting this optimum combination from the measuring instrument 70 on which the articles are placed. The removed articles are collected as one R1 rank combination product and sent to the next process for shipping, for example.

  Similarly, when the optimal combination is selected for, for example, the R2 rank, the control device 30 gives, for example, a red color “R2” to each measuring instrument 70 on which the articles constituting the optimal combination are placed. An instruction is sent to turn on the lamp 86. As a result, the “R2” lamp 86 of the measuring instrument 70 that has received this instruction is lit, and the operator has selected the optimum combination for the R2 rank, and the “R2” lamp 86 is lit. Recognizing that the articles placed on the weighing instrument 70 constitute an optimum combination of R2 ranks. Then, the operator manually removes the articles constituting this optimum combination from the measuring instrument 70 on which the articles are placed. The removed articles are collected together as a combination product of R2 rank and sent to the next process.

  Further, when the optimum combination is selected for the R3 rank, the control device 30 provides, for example, a blue lamp 88 “R3” to each measuring instrument 70 on which the articles constituting the optimum combination are placed. Send instructions to light up. As a result, the “R3” lamp 88 of the weighing instrument 70 that has received this instruction is turned on, and the operator removes the article from the weighing instrument 70 for which this R3 ”lamp 88 is lit. Are combined into one product and sent to the next process.

  The control device 30 sends an instruction for turning off the lit lamps 84, 86, or 88 to the measuring instrument 70 that has become empty due to the removal of the article. As a result, the lamp 84, 86 or 88 of the measuring instrument 70 that has received this instruction is turned off. Then, when a new article is placed on the empty measuring instrument 70, the control device 30 performs a combination operation again in the same manner as described above.

  By the way, since the articles | goods in this embodiment are primary products called agricultural products, as shown in FIG. 4, it can be considered that it distributes substantially as a weight value wx as a variable. Note that the weight distribution of an article having a substantially normal distribution is determined by, for example, the average value wm of the weight value wx of the article and the standard deviation σ. For example, prior to the combination weighing operation, the average weight value wm and the standard deviation σ of the articles are obtained by calculating the weight values wx of a number of articles of the same type to be subjected to the combination weighing operation. Alternatively, it can be obtained by actual measurement using another weighing device. The product ranks R1, R2, and R3 described above are set as follows with respect to the weight distribution of articles that are substantially normally distributed as described above.

  That is, first, apart from the product ranks R1, R2 and R3, the article weight ranks (hereinafter referred to as article ranks) as five small regions r1, r2, r3, r4 and r5 are set. Specifically, w1 = wm−3 · σ, w2 ′ = wm−1.3 · σ, w2 = wm−0.3 · σ, w3 = wm + 0.3 · σ, w3 ′ = wm + 1.3 · σ And six boundary weight values of w4 = wm + 3 · σ are set, and an area of w1 ≦ wx <w2 ′ among them is set as the r1 rank. Then, the region of w2 ′ ≦ wx <w2 is set as the r2 rank, and the region of w2 ≦ wx <w3 is set as the r3 rank. Further, the region of w3 ≦ wx <w3 ′ is set as r4 rank, and the region of w3 ′ ≦ wx ≦ w4 is set as r5 rank. Needless to say, the r1 rank and the r5 rank, the r2 rank and the r4 rank are symmetric with respect to the average weight value wm, and the r3 rank itself is also symmetric with respect to the average weight value wm.

  The product ranks R1, R2, R3, and R3 are configured by combining these article ranks r1, r2, r3, r4, and r5. Specifically, the R1 rank is composed of the r1 rank and the r2 rank. That is, articles having a weight value wx belonging to either the r1 rank or the r2 rank are selected as the R1 rank. The r2 rank, the r3 rank, and the r4 rank constitute the R2 rank. That is, an article having a weight value wx belonging to any one of the r2 rank, the r3 rank, and the r4 rank is selected as the R2 rank. Furthermore, the R3 rank is composed of the r4 rank and the r5 rank. That is, articles having a weight value wx belonging to either the r4 rank or the r5 rank are selected as the R3 rank. It is obvious that the R1 rank and the R3 rank are symmetric with respect to the average weight value wm, and that the R2 rank itself is also symmetric with respect to the average weight value wm.

  With such a setting, the weight value widths Δw1 (= w2−w1), Δw2 (= w3′−w2 ′) and Δw3 (= w4−w3) of the product ranks R1, R2 and R3 are substantially equivalent to each other. . That is, the weight value width Δw1 of the R1 rank is Δw1 = 2.7 · σ, the weight value width Δw2 of the R2 rank is Δw2 = 2.6 · σ, and the weight value width Δw3 of the R3 rank is Δw3 = 2.7 · σ. σ. Accordingly, the variation in the weight value wx of the articles belonging to each of the product ranks R1, R2, and R3 is substantially uniform, and hence the variation in the weight value w of the articles constituting the optimum combination is also substantially uniform. .

  Of the article ranks r1, r2, r3, r4, and r5, the r2 rank is included in both the R1 rank and the R2 rank, but the articles belonging to the r2 rank as the overlapping area are 80 [%] is sorted into the R1 rank, and the remaining 20 [%] is sorted into the R2 rank. The r4 rank is included in both the R2 rank and the R3 rank, and 20% of the articles belonging to the r4 rank are selected as the R2 rank, and the remaining 80 [%] is R3. Sorted into ranks. By performing such selection, the distribution ratios p1, p2, and p3 of articles belonging to each of the product ranks R1, R2, and R3 are made substantially uniform.

  More specifically, it is assumed that, for example, the distribution ratio of the entire article is 1. In this case, the distribution ratio pa of articles belonging to the r1 rank is expressed by the following formula 5 from the standard normal distribution table shown in FIG.

<< Formula 5 >>
pa = p [w1] −p [w2 ′] = 0.4987−0.4032 = 0.0955

  Then, the distribution ratio pb of articles belonging to the r2 rank is expressed by the following formula 6.

<< Formula 6 >>
pb = p [w2 ′] + p [w2] = 0.4032-0.1179 = 0.2853

  Further, the distribution ratio pc of articles belonging to the r3 rank is as shown in the following Expression 7.

<< Formula 7 >>
pc = p [w2] + p [w3] = 0.179 + 0.1179 = 0.2358

  Here, since the R1 rank is composed of the r1 rank and the r2 rank, and 80% of the articles belonging to the r2 rank are selected as the R1 rank, the distribution of the articles belonging to the R1 rank. The ratio p1 is expressed by the following formula 8.

<< Formula 8 >>
p1 = pa + 0.8 · pb = 0.0955 + 0.8 × 0.2853 = 0.3237

  For the distribution ratio p2 of articles belonging to the R2 rank, the R2 rank is composed of the r2 rank, the r3 rank, and the r4 rank, and 20% of the articles belonging to the r2 rank are selected as the R2 rank. Of the articles belonging to the r4 rank that is symmetrical to the r2 rank, 20% is selected as the R2 rank, so that the following Expression 9 is obtained.

<< Formula 9 >>
p2 = {0.2 · pb} × 2 + pc = {0.2 × 0.2853} × 2 + 0.2358 = 0.3499

  Further, the distribution ratio p3 of the articles belonging to the R3 rank is “0.3237” which is the same as the distribution ratio p1 of the articles belonging to the R1 rank because the R3 rank is in a symmetrical relationship with the R1 rank.

  In this way, the distribution ratio p1 of articles belonging to the R1 rank and the distribution ratio p3 of articles belonging to the R3 rank are each “0.3237”, and the distribution ratio p2 of articles belonging to the R2 rank is also “0. 3499 "and substantially the same value. As a result, the optimum combination constituted by the articles belonging to the R1 rank and the optimum combination constituted by the articles belonging to the R3 rank are each generated at a ratio of 32.37 [%], and are constituted by the articles belonging to the R2 rank. It is expected that the optimal combination to be generated is also generated at approximately the same ratio of 34.99 [%]. That is, the combination products of the product ranks R1, R2 and R3 can be generated with a uniform ratio and with a good balance.

  As described above, the articles belonging to the r2 rank as the overlapping region are sorted into the R1 rank and the R2 rank at a ratio of 4: 1 (= 80: 20), and the same ratio is also applied to the articles belonging to the r4 rank. In order to realize this, the control device 30 (CPU 36) includes an addition register as shown in FIG.

  That is, when the optimal combination of the R1 rank is generated by the articles belonging to the r2 rank, the control device 30 counts the number a1 of articles belonging to the r2 rank among the articles constituting the optimal combination. Then, the count value a1 is sequentially added to the addition register of FIG. In addition, even when the optimal combination of R2 rank is generated by the articles belonging to the r2 rank, the number a2 of articles belonging to the r2 rank among the articles constituting the optimal combination is counted, and the count value a2 Add sequentially to the addition register. Further, when the R2 rank optimum combination is generated by the article belonging to the r4 rank, the number b2 of articles belonging to the r4 rank among the articles constituting the optimum combination is counted, and this count value b2 is stored in the addition register. Add sequentially. Even when an R3 rank optimum combination is generated by an article belonging to the r4 rank, the number b3 of articles belonging to the r4 rank among the articles constituting the optimum combination is counted, and this count value b3 is added. Add sequentially to the register.

  In the process in which the count values a1, a2, b2, and b3 are sequentially added in this manner, for example, when a1 = 0 or {a2 / a1} ≧ ¼, the control device 30 determines that r2 The combination calculation is performed after the articles belonging to the rank are preferentially sorted into the R1 rank. That is, for the R1 rank, the optimum combination is selected from the articles belonging to the r1 rank and the r2 rank. For the R2 rank, an optimal combination is selected by excluding at least articles belonging to the r2 rank. However, when the optimum combination is not selected for the R1 rank, the articles belonging to the r2 rank are sorted into the R2 rank, and then the optimum combination is selected for the R2 rank again. That is, the highest priority is given to the selection of the optimum combination regardless of the R1 rank or the R2 rank.

  On the other hand, when {a2 / a1} <1/4, the control device 30 preferentially selects articles belonging to the r2 rank as the R2 rank, and then performs a combination operation. That is, for the R1 rank, the optimal combination is selected from the articles belonging only to the r1 rank by excluding the articles belonging to the r2 rank. For the R2 rank, the optimal combination including the articles belonging to the r2 rank is selected. Do. However, in this case as well, if the optimum combination is not selected for the R2 rank, the articles belonging to the r2 rank are sorted into the R1 rank, and then the optimum combination for the R1 rank is selected again.

  Further, when b3 = 0 or {b2 / b3} ≧ ¼, the control device 30 preferentially selects articles belonging to the r4 rank as the R3 rank, and then performs a combination operation. That is, for the R3 rank, the optimum combination is selected from the articles belonging to the r4 rank and the r5 rank. For the R2 rank, the optimum combination is selected excluding at least articles belonging to the r4 rank. However, when the optimal combination is not selected for the R3 rank, the articles belonging to the r4 rank are selected as the R2 rank, and then the optimal combination is selected for the R2 rank again.

  When {b2 / b3} <1/4, the control device 30 preferentially selects articles belonging to the r4 rank as the R2 rank, and then performs a combination operation. That is, for the R3 rank, the optimal combination is selected from the articles belonging only to the r5 rank by excluding the articles belonging to the r4 rank, and for the R2 rank, the optimal combination including the articles belonging to the r4 rank is selected. Do. However, also in this case, if the optimal combination is not selected for the R2 rank, the articles belonging to the r4 rank are sorted into the R3 rank, and then the optimal combination is selected for the R3 rank.

  As described above, according to the present embodiment, the variation in the weight value wx of the articles belonging to each of the product ranks R1, R2, and R3 is substantially uniform, and thus the weight value w of the articles constituting the optimum combination. Also, the variation is substantially uniform. In addition, the distribution ratios p1, p2 and p3 of the articles belonging to each of the product ranks R1, R2 and R3 are made uniform, so that the optimum combination of the product ranks R1, R2 and R3 is generated at a uniform ratio. . Therefore, the variation of the weight value wx of the articles constituting the optimal combination can be suppressed uniformly in any of the product ranks R1, R2 and R3, and the combination products of the product ranks R1, R2 and R3 can be uniformly distributed. It can be generated with a good balance.

  In the present embodiment, relatively large agricultural products such as apples and onions are exemplified as articles, but the present invention is not limited thereto. For example, other agricultural products such as vegetables and fruits may be used, and marine products such as seafood and seaweeds may be used. And not only these primary products such as agricultural products and marine products, but also processed products such as broiler (processed meat) and kamaboko, so-called secondary products may be used.

  Further, although the allowable weight range is defined for each of the product ranks R1, R2, and R3 with the target weight values Wt1, Wt2, and Wt3 as lower limits, the present invention is not limited to this. That is, the reason why the respective allowable weight ranges are defined with the target weight values Wt1, Wt2, and Wt3 as lower limits is to guarantee the minimum weight value of the product. For example, the target weight value is set at the center of each allowable weight range. What is necessary is just to set it as the aspect according to each use, such as setting only target weight value Wt1, Wt2, and Wt3, without setting Wt1, Wt2, and Wt3, or providing an allowable weight range.

  For reference, for example, for the R2 rank in the center, if the target weight value Wt2 is set so that Wt2 = wm · M2, the optimum combination can be easily selected, and the weighing instrument 70 (the weighing platform 72) will remain on forever. It is possible to suppress the generation of a so-called staying product that remains in the state of being placed. For the R1 rank, the deviation from the average weight value Wt1 / M1 to the weight value wx of each article belonging to the R1 rank is determined based on the average weight value Wt1 / M1 per article constituting the optimum combination. If the target weight value Wt1 is set so that the sum is substantially zero, the optimum combination is easily selected. However, in this case, the r2 rank constituting the R1 rank is calculated assuming that there are only 80 [%] articles assigned to the R1 rank. For the R3 rank, it is desirable to set the target weight value Wt3 in the same manner as the R1 rank.

  Furthermore, in this embodiment, the three lamps 84, 86, and 88 corresponding to the product ranks R1, R2, and R3 are provided in each weighing instrument 70, but the present invention is not limited to this. For example, these lamps 84, 86, and 88 may be combined into one by using a light emitting element that can be changed to an arbitrary emission color such as a generally known RGB-LED (light emitting diode). Further, even in the case of a single color light emitting element, the same operation as that provided by the plurality of lamps 84, 86 and 88 can be achieved by changing the lighting mode (flashing or changing the flashing cycle). You may do it.

  Further, in the present embodiment, in a state where the articles are placed on the respective weighing instruments 70, the articles are classified according to the product ranks R1, R2, and R3 (strictly, the article ranks r1, r2, r3, r4, and r5). In addition, the combination calculation for selecting the optimum combination is performed, that is, the selection operation and the combination weighing operation are performed in parallel. However, the present invention is not limited to this. For example, similar to the above-described prior art, prior to the combination weighing operation, the selection operation may be performed first, and then the combination weighing operation may be performed for each of the product ranks R1, R2, and R3. In an extreme case, a sorting device for performing sorting work may be provided separately from the combination weigher for performing combination weighing work. That is, after the sorting operation is performed by the sorting device, the combination weighing operation may be performed by a combination scale different from the sorting device.

  Further, the number of each article rank r1, r2, r3, r4 and r5, the value of each boundary weight value w1, w2 ′, w2, w3, w3 ′ and w4, the r2 rank and the r4 rank as overlapping regions, respectively. The selection ratio of the articles is not limited to the aspect described in the present embodiment. In short, the weight value widths Δw1, Δw2 and Δw3 of each of the product ranks R1, R2 and R3 are substantially equivalent to each other, and the distribution ratio of articles in each of the product ranks R1, R2 and R3 is substantially uniform. What is necessary is just to determine these aspects suitably.

  In this embodiment, three product ranks R1, R2, and R3 are set, but the present invention is not limited to this. For example, a case where four product ranks R1, R2, R3, and R4 are set will be described with reference to FIG.

  That is, first, w1 = wm−3 · σ, w2 ′ = wm−2.5 · σ, w2 = wm−0.5 · σ, w3 = wm + 0.5 · σ, w3 ′ = wm + 2.5 · σ and Six boundary weight values of w4 = wm + 3 · σ are set. In addition, the average weight value wm of the article is also used as one boundary weight value. Then, six article ranks r1, r2, r3, r4, r5 and r6 are set by these seven boundary weight values w1, w2 ', w2, wm, w3, w3' and w4. Specifically, the region of w1 ≦ wx <w2 ′ is set as the r1 rank, the region of w2 ′ ≦ wx <w2 is set as the r2 rank, and the region of w2 ≦ wx <wm is set as the r3 rank. Then, the region of wm ≦ wx <w3 is set as r4 rank, the region of w3 ≦ wx <w3 ′ is set as r5 rank, and the region of w3 ′ ≦ wx ≦ w4 is set as r6 rank. Needless to say, the r1 rank and the r6 rank, the r2 rank and the r5 rank, and the r3 rank and the r4 rank are respectively symmetrical with respect to the average weight value wm.

  Then, four product ranks R1, R2, R3 and R4 are configured by combining these six article ranks r1, r2, r3, r4, r5 and r6. Specifically, the R1 rank is configured by the r1 rank and the r2 rank, and the R2 rank is configured by the r2 and r3 rank. The r4 rank and the r5 rank constitute the R3 rank, and the r5 rank and the r6 rank constitute the R4 rank. Needless to say, the R1 rank and the R4 rank, and the R2 rank and the R3 rank are respectively symmetrical with respect to the average weight value wm.

  By such setting, the weight value widths Δw1 (= w2−w1), Δw2 (= wm−w2 ′), Δw3 (= w3′−wm) and Δw4 (= w4) of the product ranks R1, R2, R3 and R4, respectively. -W3) is 2.5 · σ in both cases. That is, in each of these product ranks R1, R2, R3, and R4, the variation in the weight value wx of the articles belonging to each of them is 2.5 · σ or less. Therefore, the weight value w of the articles constituting the optimal combination Variations can be suppressed uniformly.

  For articles belonging to the r2 rank that are included in both the R1 rank and the R2 rank, 80 [%] is selected as the R1 rank, and the remaining 20 [%] is selected as the R2 rank. . Also, for articles belonging to the r5 rank that are included in both the R3 rank and the R4 rank, 20 [%] is selected as the R3 rank, and the remaining 80 [%] is selected as the R4 rank. . By doing so, the distribution ratios p1, p2, p3, and p4 of articles belonging to each of the product ranks R1, R2, R3, and R4 are made substantially uniform.

  More specifically, it is assumed that, for example, the distribution ratio of the entire article is 1. In this case, the distribution ratio pa of articles belonging to the r1 rank is expressed by the following formula 10 from the standard normal distribution table shown in FIG.

<< Formula 10 >>
pa = p [w1] −p [w2 ′] = 0.4987−0.4938 = 0.499

  Then, the distribution ratio pb of articles belonging to the r2 rank is as shown in the following Expression 11.

<< Formula 11 >>
pb = p [w2 ′] + p [w2] = 0.40938−0.1915 = 0.0303

  Further, the distribution ratio pc of articles belonging to the r3 rank is expressed by the following Expression 12.

<< Formula 12 >>
pc = p [w2] -0 = 0.915

  Here, since the R1 rank is composed of the r1 rank and the r2 rank, and 80% of the articles belonging to the r2 rank are selected as the R1 rank, the distribution ratio of the articles belonging to the R1 rank. p1 is expressed by the following equation (13).

<< Formula 13 >>
p1 = pa + 0.8 · pb = 0.499 + 0.8 × 0.3023 = 0.2467

  As for the distribution ratio p2 of the articles belonging to the R2 rank, the R2 rank is composed of the r2 rank and the r3 rank, and 20 [%] of the articles belonging to the r2 rank is selected as the R2 rank. The following equation 14 is obtained.

<< Formula 14 >>
p2 = 0.2 · pb + pc = 0.2 × 0.3023 + 0.1915 = 0.2520

  Note that the distribution ratio p3 of articles belonging to the R3 rank is “0.2520”, which is the same as the distribution ratio p2 of articles belonging to the R2 rank, and the distribution ratio p4 of articles belonging to the R4 rank is the article belonging to the R1 rank. It is clear that the distribution ratio p1 is “0.2467”.

  As described above, the distribution ratio p1 of articles belonging to the R1 rank and the distribution ratio p4 of articles belonging to the R4 rank are “0.2467”, respectively, and the distribution ratio p2 of articles belonging to the R2 rank and the R3 rank belong to the R3 rank. The article distribution ratio p3 is also substantially the same value of “0.2520”. Therefore, the combination products of the product ranks R1, R2, R3, and R4 can be generated in a uniform ratio with a good balance.

  In the present embodiment, the manual combination balance 10 has been described as an example, but the present invention can be applied to a semi-automatic or automatic combination balance. For reference, the semi-automatic combination weigher is a method that is manually performed when individual items are supplied to each measuring instrument and automatically performed when individual items are removed from the measuring instrument. It is. The automatic method is a method in which supply and removal of individual products are both performed automatically.

It is a block diagram which shows schematic structure of one Embodiment of this invention. It is a block diagram which shows the detail of the control apparatus in the embodiment. It is a block diagram which shows the detail of each measuring device in the embodiment. It is an illustration figure which shows the relationship between the weight distribution of the articles | goods in the embodiment, and each product rank etc. set to this. It is an illustration figure which shows notionally the structure of the addition register with which the control apparatus in the embodiment is provided. It is an illustration figure which shows another example of the embodiment. It is an illustration figure for demonstrating the problem in a prior art. It is a figure which shows the standard normal distribution table generally known. It is an illustration figure different from FIG. 7 for demonstrating the problem in a prior art.

Explanation of symbols

10 Combination Weigher 30 Controller 36 CPU
70 Weighing scale

Claims (7)

Sorting means for sorting articles having a substantially normal distribution with a weight value as a variable according to a plurality of weight ranks according to the weight value;
Combination selection means for selecting various combinations of weight values of the articles belonging to the same weight rank and selecting a combination having a combination weight value equal to or closest to the target weight value;
Comprising
The weight value widths of each of the plurality of weight ranks are substantially equal to each other, and each of the plurality of weight ranks has an overlapping region overlapping with the other weight ranks.
Combination scale. Each of the plurality of weight ranks consists of a plurality of small regions partitioned by a plurality of boundary weight values,
Some of the small areas are shared by the two or more weight ranks as the overlapping area.
The combination weigher according to claim 1. The sorting means sorts the articles belonging to the overlapping area into the weight ranks including the overlapping area at a predetermined sorting ratio;
The combination weigher according to claim 1 or 2. The selection ratio is determined so that the distribution ratio of the articles belonging to each of the plurality of weight ranks is substantially equal between the plurality of weight ranks.
The combination weigher according to claim 3. A plurality of weighing means for measuring the weight value of each of the articles supplied and the articles supplied;
In a state where the article is supplied to each of the plurality of weighing units, the selection by the selection unit and the selection by the combination selection unit are continuously performed based on the weight measurement value by each of the plurality of weighing units. ,
The combination weigher according to any one of claims 1 to 4. It has a sorting mode and a combination mode,
Performing selection by the combination selection unit by the combination mode after performing selection by the selection unit by the selection mode;
The combination weigher according to any one of claims 1 to 4. In a sorting apparatus that sorts articles having a substantially normal distribution with a weight value as a variable according to a plurality of weight ranks according to the weight value,
Used in combination weighers that combine various weight values of the articles belonging to the same weight rank to select a combination whose combined weight value is equal to or closest to the target weight value,
The weight value widths of the plurality of weight ranks are substantially equal to each other, and each of the plurality of weight ranks has an overlapping region overlapping with the other weight ranks;
A sorting apparatus for a combination weigher.

Images