Combination scales

Abstract

We provide a combination scale that can improve combination accuracy. [Solution] A combination scale comprising a plurality of hoppers (weighing hopper 70 and memory hopper 80) for containing the objects to be weighed, and a control device 110 that performs combination calculations based on the weights of the objects to be weighed contained in the plurality of hoppers (weighing hopper 70 and memory hopper 80), wherein the control device 110 acquires a plurality of (e.g., N) combination patterns in which the combined weight of the objects to be weighed falls within a predetermined weight range, and selects one combination pattern from the plurality of combination patterns based on a first condition (any of combination selections A to F).

Description

Technical Field

[0001] The present invention relates to the technology of a combination scale that performs combination operations.

Background Art

[0002] Conventionally, the technology of a combination scale that performs combination operations has been known. For example, it is as described in Patent Document 1.

[0003] Patent Document 1 describes a combination scale that measures the weights of objects to be weighed accommodated in a plurality of weighing hoppers respectively, combines the weights of the objects to be weighed in various ways by combination operations, and selects a weighing hopper of an appropriate combination in which the total weight is within a target weight range from these combinations.

Prior Art Documents

[0009] According to the present invention, the accuracy of combinations can be improved. [Brief explanation of the drawing]

[0010] [Figure 1] A schematic diagram showing the general configuration of a combination scale according to one embodiment of the present invention. [Figure 2] A block diagram showing the configuration of a combination scale. [Figure 3] A flowchart illustrating the main control performed by the control unit during the learning phase. [Figure 4] A flowchart illustrating the combination recording control, one of the main control functions performed by the control unit. [Figure 5] A flowchart illustrating the combination selection control, one of the main control functions performed by the control unit. [Figure 6] A flowchart illustrating the first priority execution process in combination selection control. [Figure 7] A flowchart illustrating the second-priority execution process in combination selection control. [Figure 8] A flowchart illustrating the main control performed by the control device during the utilization phase. [Modes for carrying out the invention]

[0011] In the following, a combination scale 1 according to one embodiment of the present invention will be described with reference to Figures 1 and 2.

[0012] The combination scale 1 according to this embodiment determines a combination of weighed objects such that the total weight of the weighed objects falls within a predetermined weight range. In this embodiment, the combination scale 1 is automatic, but it may also be semi-automatic or manual. Various granular weighed objects (e.g., grains, dried fruits, and other foods) can be used as weighed objects. The combination scale 1 is configured to discharge the determined combination of weighed objects to a packaging machine 200, which will be described later.

[0013] The combination scale 1 comprises a supply device 10, a base 20, a distribution feeder 30, a level sensor 40, a linear feeder 50, a supply hopper 60, a weighing hopper 70, a memory hopper 80, a collection unit 90, an operation setting display device 100, and a control device 110.

[0014] The supply device 10 is capable of supplying the object to be weighed from a predetermined supply source to the distribution feeder 30. The supply device 10 includes a bucket lifter (not shown) for transporting the object to be weighed to a predetermined height position, a supply feeder 11 for transporting the object to be weighed from the bucket lifter to the tip side, and a vibration mechanism (not shown) for vibrating the supply feeder 11. In this way, the supply device 10 transports the object to be weighed to the tip side of the supply feeder 11 by vibration of the supply feeder 11, and also causes it to fall downward (to the distribution feeder 30) from the tip of the supply feeder 11.

[0015] The base 20 is located approximately in the center of the combination scale 1. The base 20 is formed in a hollow, roughly box-like shape and houses various devices inside.

[0016] The distribution feeder 30 is used to disperse the material to be weighed. The material to be weighed is supplied to the distribution feeder 30 from a predetermined supply device. The distribution feeder 30 has a top cone 31 that is roughly conical (or frustoconical) in shape, and an excitation mechanism 32 that vibrates the top cone 31. In this way, the distribution feeder 30 disperses the supplied material to be weighed radially outward along the slope of the top cone 31 by the vibration of the top cone 31, and also causes it to fall downward (to the straight feeder 50) from the radially outer end of the top cone 31.

[0017] The level sensor 40 measures the amount of the weighed object on the top cone 31. The level sensor 40 is provided above the top cone 31.

[0018] The straight feeder 50 linearly conveys the weighed object supplied from the dispersion feeder 30 radially outward of the base body 20. A plurality (14) of straight feeders 50 are arranged along the circumferential direction of the base body 20 so as to correspond to the number of heads. The straight feeder 50 includes a straight trough 51 extending along the radial direction of the base body 20 and a vibration mechanism 52 that vibrates each straight trough 51. Thus, the straight feeder 50 moves the weighed object supplied from the dispersion feeder 30 on the straight trough 51 toward the tip side by the vibration of the straight trough 51 and drops it downward (supply hopper 60) from the tip of the straight trough 51.

[0019] The supply hopper 60 temporarily stores the weighed object supplied from the straight feeder 50 and supplies the stored weighed object to the weighing hopper 70. A plurality (14) of supply hoppers 60 are arranged along the circumferential direction on the radially outer side of the base body 20 so as to correspond to each straight feeder 50.

[0020] Further, the supply hopper 60 is formed in a bottomed box shape with an open top and has a discharge gate (not shown) at the bottom that can be controlled to open and close. Thus, the supply hopper 60 drops the stored weighed object downward (weighing hopper 70) by opening the bottom.

[0021] The weighing hopper 70 measures the weight of the weighed object supplied from the supply hopper 60, temporarily stores it, and selectively supplies the stored weighed object to the collecting unit 90 or the memory hopper 80. The weighing hopper 70 has a weight sensor 71 constituted by, for example, an appropriate load cell for measuring the weight of the weighed object. The weighing hopper 7 is arranged below each supply hopper 60. That is, a plurality (14) of weighing hoppers 70 are arranged along the circumferential direction on the radially outer side of the base body 20.

[0022] The weighing hopper 70 is formed in the shape of a bottomed box with an open top and has a discharge gate (not shown) at the bottom that can be opened and closed. In this way, the weighing hopper 70 allows the stored weighing material to fall into the collection section 90 by opening a part of the gate (radially outward from the base 20). The weighing hopper 70 also allows the stored weighing material to fall into the memory hopper 80 by opening the other part of the gate (radially inward from the base 20).

[0023] The memory hopper 80 temporarily holds the material to be weighed supplied from the weighing hopper 70 and supplies the held material to the collection unit 90. The memory hopper 80 is positioned diagonally below each weighing hopper 70 (radially inward from the base 20). In other words, multiple memory hoppers (14 in total) are arranged radially outward from the base 20, along the circumferential direction.

[0024] The memory hopper 80 is formed in the shape of a bottomed box with an open top and has a gate (not shown) at the bottom that can be opened and closed for discharge. In this way, the memory hopper 80 drops the held weighing material into the collection section 90 by opening the gate.

[0025] The collection unit 90 collects the material to be weighed supplied from the weighing hopper 70 or the memory hopper 80 and is capable of loading it into the material to be weighed. The collection unit 90 has a collection hopper (not shown) that temporarily stores the material to be weighed from the weighing hopper 70. The collection unit 90 discharges the stored material to be weighed to the packaging machine 200 based on a discharge command from the packaging machine 200, for example.

[0026] The operation setting display device 100 shown in Figure 2 is capable of displaying various types of information. The operation setting display device 100 is composed of a touch panel that allows for touch operation. In other words, the operation setting display device 100 serves as both a display device for displaying various types of information and an input device for inputting various types of information. The operation setting display device 100 can be installed at any location on the combination scale 1.

[0027] The control device 110 shown in Figure 2 comprises a CPU unit 111, a memory 112, a supply device control circuit 113, an A / D conversion circuit unit 114, a feeder control circuit unit 115, a hopper discharge gate control circuit 116, an A / D conversion circuit unit 117, and an I / F circuit 118.

[0028] The CPU unit 111 controls the overall operation of the combination scale 1. The CPU unit 111 controls each part and also performs various control (processing) such as combination calculations. The memory 112 stores the operation programs for each part and the set operation parameters, and serves as a work area for calculations and other operations performed by the CPU unit 111.

[0029] The supply device control circuit 113 controls, for example, the excitation mechanism of the supply device 10 based on a control signal from the CPU unit 111. The A / D conversion circuit unit 114 converts the analog signal from the level sensor 40 into a digital signal and outputs it to the CPU unit 111. The feeder control circuit unit 115 controls the excitation mechanism 32 of the distributed feeder 30 and the excitation mechanisms 52 of each linear feeder 50 based on a control signal from the CPU unit 111. The hopper discharge gate control circuit 116 controls the opening and closing of the discharge gates for the supply hopper 60, weighing hopper 70, memory hopper 80 and collection hopper based on a control signal from the CPU unit 111. The A / D conversion circuit unit 117 converts the analog signal from the weight sensor 71 of the weighing hopper 70 into a digital signal and outputs it to the CPU unit 111. The I / F circuit 118 connects the CPU unit 111 to the operation setting display device 100.

[0030] The control device 110, configured as described above, operates based on various setting values ​​input to the operation setting display device 100. These setting values ​​include operating parameters.

[0031] For example, the control device 110 is configured with a model and type based on input to the operation setting display device 100. Here, "model" includes the type and specifications of the combination scale 1 and the equipment of each component that makes up the combination scale 1. By configuring the model in this way, the method of loading the object to be weighed, the number of heads, the capacity of each hopper, the presence or absence of a memory hopper 80, etc. are specified. The type of combination scale 1 (for example, automatic, semi-automatic, or manual) is also specified by configuring the model.

[0032] Furthermore, the "product type" setting includes a collection of operating parameters that differ depending on the type of object being weighed and the target combined weight. By switching the product type setting in this way, the various operating parameters necessary for weighing the object can be switched to those for various objects and target combined weights.

[0033] In the combination scale 1 configured as described above, the linear feeder 50, supply hopper 60, weighing hopper 70, and memory hopper 80 constitute one set of heads. In this embodiment, there are 14 sets of heads. Each head is associated with a unique number (head number) and managed by the control device 110. The supply feeder 11, supply hopper 60, weighing hopper 70, and memory hopper 80 that constitute each head constitute the main body of the combination scale (see Figure 2).

[0034] The following describes the basic operation of the combination scale 1 performed by the control device 110.

[0035] The control device 110 controls the supply device 10 to supply the material to be weighed to the distribution feeder 30. The control device 110 then operates the distribution feeder 30 to transport the material to be weighed from the distribution feeder 30 to each head (on the straight feeder 50 side). The control device 110 then operates the straight feeder 50 to transport the material to be weighed from the straight feeder 50 to the supply hopper 60.

[0036] The control device 110 then weighs the materials in each weighing hopper 70 after loading them from the supply hopper 60. If the memory hopper 80 is empty, the control device 110 loads the materials from the weighing hopper 70, which is currently storing the materials, into the memory hopper 80. In this way, each hopper (from weighing hopper 70 to memory hopper 80) is filled with materials for which the weighed value (weight value) has already been acquired.

[0037] The control device 110 also calculates a total weight value (combined weight value) by appropriately combining the weight values ​​in each weighing hopper 70 and the weight values ​​in each memory hopper 80. The control device 110 then obtains a combination of hoppers in which the total weight value falls within a predetermined weight range.

[0038] In this way, the control device 110 can perform a calculation (combination calculation) to sum the weight values ​​of the objects to be weighed contained in each hopper (weighing hopper 70 and memory hopper 80) and calculate the total weight. The predetermined weight range is a range that is greater than or equal to a preset target combined weight value, and less than or equal to the sum of the target combined weight and the allowable upper limit value.

[0039] The control device 110 then controls the hoppers (weighing hopper 70 and memory hopper 80) included in the selected combination and opens the discharge gates of the hoppers, thereby feeding the weighed items contained in the hoppers into the collection unit 90. As a result, the combination scale 1 can feed weighed items within a predetermined weight range from the hoppers (weighing hopper 70 and memory hopper 80) through the collection unit 90 to the packaging machine 200.

[0040] Conventionally, with typical combination scales, the combination accuracy may decrease as operation continues. Below, we will explain the factors that cause a decrease in combination accuracy in typical combination scales. Six factors will be given as examples, and each factor will be referred to as Factor A to F.

[0041] First, let's explain factor A. In other words, as the combination scale continues to operate, there may be hoppers into which objects weighed that are unlikely to be selected for combinations are loaded (hereinafter referred to as "stagnant hoppers"). When such stagnant hoppers exist, the combination accuracy decreases.

[0042] Next, let's explain factor B. In other words, in order to correct for fluctuations in the zero point, the weighing hopper may be emptied for a predetermined period of time to perform zero correction. In this case, the weighing hopper on which zero correction is performed is forcibly included in the combination (even if the total weight value is not within the predetermined weight range). If a specific hopper is always included in the combination in this way, the accuracy of the combination will decrease.

[0043] Next, let's explain factor C. Specifically, if a weighing hopper and a memory hopper of the same head are selected for combination at the same time, the number of hoppers that can participate in the combination (as candidates) in the next weighing cycle decreases, thus reducing the combination accuracy.

[0044] Next, let's explain factor D. That is, if a large number of hoppers are selected for a combination, the number of hoppers that can participate in the next weighing cycle decreases, thus reducing the combination accuracy.

[0045] Next, let's explain factor E. In the case of a combination scale, if it is a semi-automatic combination scale, the weighing hopper is a two-chamber type. Therefore, when both weighing chambers become empty, the number of hoppers that can participate in the combination in the next weighing cycle decreases, resulting in a decrease in combination accuracy.

[0046] Next, let's explain factor F. Specifically, if the combination scale is a semi-automatic type, the operator must move from one end of the machine to the other. In this case, the increased momentum of the operator becomes a burden, leading to a decrease in calculation capacity. As a result, the combination accuracy decreases.

[0047] To improve the accuracy of combinations for these factors A to F, the applicant devised a method (hereinafter referred to as "combination selection") that involves preparing multiple combinations in which the total weight value falls within a predetermined weight range, and using a combination appropriately selected from these. The combination selection for factors A to F will be explained below. Hereafter, the combination selection for each of factors A to F will be referred to as combination selection A to F, respectively.

[0048] First, let's explain combination selection A for factor A. That is, if a stagnant hopper occurs, we select a combination that includes the stagnant hopper from among several combinations. In other words, we prioritize including the stagnant hopper in the combination over other hoppers. In this way, the stagnant hopper can be eliminated, and the accuracy of the combination is improved.

[0049] Next, we will explain combination selection B for factor B. That is, since zero correction is performed at predetermined intervals, when the predetermined period approaches, a combination that includes the head to be zero corrected (i.e., the weighing hopper) is selected from multiple combinations, the weighing hopper is emptied, and the zero point is calibrated. In this way, the weighing hopper is not forcibly included in the combination under undesirable circumstances, and the combination accuracy is improved.

[0050] Next, we will explain combination selection C for factor C. That is, from multiple combinations, we select a combination that does not include weighing hoppers and memory hoppers of the same head. In this way, the number of hoppers that can participate in the combination in the next weighing cycle does not decrease, and the combination accuracy is improved.

[0051] Next, we will explain combination selection D for factor D. That is, based on the selected head number, etc., the number of hoppers that are valid (candidates for combination) in the next weighing cycle is calculated. Then, from the multiple combinations, the combination with the largest number of hoppers valid in the next weighing cycle is selected. In this way, the number of hoppers that can participate in the combination in the next weighing cycle does not decrease, and the combination accuracy is improved.

[0052] Next, we will explain combination selection E for factor E. That is, we select a combination from multiple combinations in which neither weighing tank of the weighing hopper becomes empty. In this way, the number of hoppers that can participate in the combination in the next weighing cycle does not decrease, and the combination accuracy is improved.

[0053] Next, we will explain the combination selection F for factor F. That is, from multiple combinations, we prioritize selecting combinations that include the hopper of the head near the center of the machine where the feeder is standing. In this way, the increase in the feeder's momentum is suppressed and the decrease in computational power is suppressed, thus improving combination accuracy.

[0054] Although combination selections A to F have been devised for factors A to F, the decrease in combination accuracy is also influenced by various conditions other than factors A to F, such as the model of the combination scale, the properties of the objects being weighed, and the target combination weight value, making it difficult to implement appropriate processing. In other words, if the combination accuracy decreases as the operation of combination scale 1 continues, the method of deciding which combination selection to implement becomes a problem.

[0055] Therefore, in the combination scale 1 according to this embodiment, operation is first performed over a predetermined period, and based on the results of this operation, a learning model capable of improving the combination accuracy in the execution of combination selection is generated by machine learning. In other words, the combination accuracy is improved by executing this generated learning model.

[0056] Here, machine learning refers to the process of analyzing training data, learning the patterns of features extracted from the training data, and creating a model that enables appropriate processing. As a machine learning method, AI (artificial intelligence) technologies can be employed. In this embodiment, the machine learning method employed is learning using a neural network (e.g., a convolutional neural network), which is an information processing model that mimics the structure of the neural circuits in the human brain. However, the machine learning method is not limited to the example described above; various methods such as decision trees and random forests can also be employed.

[0057] In the following, the phase in which the control device 110 of the combination scale 1 performs machine learning to generate a learning model will be referred to as the "learning phase," and the phase in which the control device 110 uses the learning model to operate the combination scale 1 will be referred to as the "utilization phase." The control of the control device 110 will be explained separately for the learning phase and the utilization phase.

[0058] First, let's explain the learning phase. During the learning phase, the control device 110 performs the main control.

[0059] In the following section, the main control performed by the control device 110 during the learning phase will be explained using the flowchart shown in Figure 3.

[0060] For example, when the operator turns on the power to the combination scale 1, the control device 110 performs a variety selection (step S1). Variety selection is the setting of the variety of the object to be weighed. Variety selection is performed, for example, by inputting into the operation setting display device 100 by the operator. Specifically, the control device 110 displays a menu screen on the operation setting display device 100. The menu screen includes a variety selection screen. The operator performs touch operations according to the contents displayed on the variety selection screen to select, for example, a variety number assigned to each variety (type) of object to be weighed. This sets the variety of the object to be weighed in the control device 110.

[0061] Next, the control device 110 sets the operating parameters (step S2). That is, the control device 110 reads and sets the various operating parameters necessary for weighing for the type of object to be weighed, which was set in step S1.

[0062] Next, the control device 110 determines whether or not to change the combination selection and changes the combination selection as necessary (steps S3, S4). Specifically, the control device 110 has pre-set priority levels for which combination selection to use for combination selections A to F. For example, combination selections A to F are assigned priority levels from 1st priority to 6th priority in order from A to F. Then, in step S4, the control device 110 shifts the priority level of each combination selection A to F by one, for example, at predetermined intervals or after a predetermined number of operations of the combination scale 1. Note that the change in combination selection may also be made, for example, by inputting into the operation setting display device 100 by the operator.

[0063] Next, the control device 110 determines whether or not the combination scale 1 is turned ON (step S5). If the control device 110 is not turned ON, it proceeds back to step S3. If the control device 110 is turned ON, it performs supply control to supply the weighing object from the supply device 10 to the distribution feeder 30 (step S6).

[0064] Next, the control device 110 performs a control (hereinafter referred to as "combination storage control") to store multiple (N, as described later) combinations of hoppers (weight values) whose total weight value falls within a predetermined weight range (step S7). Next, the control device 110 performs a control (hereinafter referred to as "combination selection control") to determine one combination from the multiple combinations stored in the combination storage control using one of combination selections A to F (step S8). A detailed explanation of combination storage control and combination selection control will be given later.

[0065] Next, the control device 110 determines whether or not there is a discharge command signal from the packaging machine 200 (step S9). If there is no discharge command signal, the control device 110 returns to step S6. If there is a discharge command signal, the control device 110 performs hopper control, which opens the discharge gates of the hoppers (weighing hopper 70 and memory hopper 80) included in the combination determined by the combination selection control (step S10). Once the hopper control is performed, the weighed items are discharged from the hoppers to the collection unit 90, and the combination calculation for one weighing cycle is completed.

[0066] Next, the control device 110 displays or transmits data (operation data) related to the processing from step S6 to step S10 to the operation setting display device 100 or other external devices (step S11). The control device 110 also stores the operation data in a predetermined storage unit. The operation data includes, for example, the determined combination selection, the combined discharge weight value discharged from the hopper to the collection unit 90, and the operating speed.

[0067] Next, the control device 110 determines whether the combination scale 1 has finished operation (is turned OFF) or not (step S12). If the control device 110 is not turned OFF, it proceeds back to step S6. If the control device 110 is turned OFF, it proceeds back to step S1.

[0068] In the following section, the combination recording control (control in step S7 of Figure 3) performed by the control device 110 will be explained using the flowchart in Figure 4.

[0069] Combination recording control is a control method that stores multiple (N) combinations of hoppers (weight values) whose total weight value falls within a predetermined weight range, as described above. N is a natural number greater than or equal to 2; for example, in this embodiment, it is 20. Furthermore, in the following, a pattern of hopper (weight value) combinations may be referred to as a "combination pattern."

[0070] First, in step S50, the control device 110 initializes a predetermined memory area (hereinafter referred to as "area MN"). In this way, the control device 110 erases the combination patterns and the combined weight values ​​of each combination pattern that were stored in the previous combination recording control. Next, in step S51, the control device 110 sets the initial combination patterns. Here, the initial combination patterns are the combination patterns of each hopper (from weighing hopper 70 to memory hopper 80), and multiple patterns are set in advance.

[0071] Next, the control device 110 executes the processes from step S52 to step S60 for all initial combination patterns. That is, the control device 110 repeats the processes from step S52 to step S60 as many times as there are initial combination patterns. In the following explanation, an initial combination pattern on which the processes from step S52 to step S60 are executed may be simply referred to as a "combination pattern".

[0072] In step S52, the control device 110 determines whether the weighing hopper 70 included in the combination pattern has completed weighing. If weighing is not completed, the control device 110 proceeds to step S61. On the other hand, if weighing is completed, the control device 110 sums the weighing values ​​of the hoppers included in this combination pattern and obtains the sum value CW (step S53).

[0073] Next, in step S54, the control device 110 determines whether the sum CW is within a predetermined weight range. If the sum CW is not within the predetermined weight range, the control device 110 proceeds to step S61. On the other hand, if the sum CW is within the predetermined weight range, the control device 110 sets n to 1 (step S55).

[0074] Here, n is a natural number greater than or equal to 1, and is associated with a combination pattern. In the multiple combination patterns stored in area MN, n is set so that the value closest to the target combination weight is 1, and the value increases as the weight value is further away. In other words, in the combination patterns stored in area MN, n indicates the order from those whose weight value is closest to the target combination weight among all combination patterns currently stored in area MN. In other words, it indicates the priority of selection for the combination patterns stored in area MN, with smaller values ​​indicating a higher likelihood of selection.

[0075] Thus, in step S54, combination patterns in which the sum CW is determined to fall within a predetermined weight range are associated with n = 1 in step S55.

[0076] Next, in step S56, the control device 110 determines whether the sum value CW obtained in step S54 is closer to the target combined weight value than the weight value of the nth combination pattern in area MN. If the sum value CW is close to the target combined weight value, the control device 110 proceeds to step S59. On the other hand, if the sum value CW is not close to the target combined weight value, the control device 110 proceeds to step S57. Here, if the sum value CW is not close to the target combined weight value (NO in step S56), it indicates that the nth combination pattern in area MN is ranked higher than the combination pattern that results in the sum value CW.

[0077] Therefore, in step S57, the control device 110 sets n, which is associated with the combination pattern that results in a combined value of CW, to a number that is n plus 1 (a new n). This sets the order of the combination pattern that results in a combined value of CW to be after the nth combination pattern of area MN.

[0078] Then, in step S58, the control device 110 determines whether n (the new n) is the same as N+1. In this embodiment, N (the upper limit of the number of combination patterns to be stored in area MN) is set to 20. In other words, in step S58, it is determined whether the number of combination patterns to be stored in area MN has reached the upper limit.

[0079] In step S58, if n (the new n) is not the same as N+1, the control device 110 proceeds to step S56 again. On the other hand, if n (the new n) is the same as N+1 (i.e., if the order of combination patterns that result in a sum value of CW exceeds the upper limit of the number of combination patterns stored in area MN), the control device 110 proceeds to step S61.

[0080] Furthermore, if the summation value CW is closer to the target combination weight value than the weight value of the nth combination pattern in area MN (YES in step S56), the control device 110 shifts the combination data from the nth to the (N-1)th combination in area MN and sets them from the (n+1)th to the Nth combination (step S59). Then, the control device 110 sets the summation value CW and its combination pattern as the nth combination (step S60).

[0081] Next, in step S61, the control device 110 determines whether the combination pattern to be processed from step S52 to step S60 is the final one. That is, the control device 110 determines whether the processing from step S52 to step S60 has been performed for all initial combination patterns.

[0082] Then, if the combination pattern is the last one, the control device 110 terminates the combination recording control. On the other hand, if the combination pattern is not the last one, the control device 110 advances the combination pattern (step S62). In this way, the control device 110 executes the process from step S52 to step S60 for the next combination pattern.

[0083] Through this combination recording control, multiple combination patterns are stored in area MN in order from the one closest to the target combination weight value (the first one). In this embodiment, the number of combination patterns stored (N) is 20, but it may be less than 20, for example, if the allowable upper limit is small.

[0084] In the following section, the combination selection control (control in step S8 of Figure 3) performed by the control device 110 will be explained using the flowchart in Figure 5.

[0085] Combination selection control is a control method that, as described above, determines one combination pattern from multiple combination patterns stored in combination memory control using one of combination selections A to F.

[0086] The control device 110 selects the first combination pattern from among the N combination patterns stored in area MN (step S100). Note that this first combination pattern is a provisional selection and may differ from the final combination pattern. After step S100, the control device 110 then performs control according to the pre-set priority order for combination selection.

[0087] By provisionally selecting the first combination pattern in this way, even if there are no combination patterns that satisfy the conditions as the final combination pattern in subsequent processing (the first priority execution process to the sixth priority execution process described later), it is possible to avoid a situation where there is no determined combination pattern.

[0088] Specifically, after step S100, the control device 110 executes the first priority execution process (step S200), the second priority execution process (step S300), the third priority execution process (step S400), the fourth priority execution process (step S500), the fifth priority execution process (step S600), and the sixth priority execution process (step S700) in order as needed (more specifically, until the final combination pattern is determined).

[0089] The following describes the first priority execution process performed by the control device 110, using the flowchart in Figure 6.

[0090] First, in step S201, the control device 110 determines whether the first priority (first priority) for combination selection is combination selection A. If the first priority is not combination selection A, the control device 110 proceeds to step S211. On the other hand, if the first priority is combination selection A, the control device 110 determines whether there are any stagnant hoppers (stagnant weighing hoppers 70 and memory hoppers 80) (step S202).

[0091] Then, if there is no stagnant hopper, the control device 110 terminates the first priority execution process and then starts the second priority execution process (step S300). On the other hand, if there is a stagnant hopper, the control device 110 executes combination selection A. That is, the control device 110 sets the combination pattern that includes the stagnant hopper and has the highest priority among the N combination patterns stored in area MN as the final combination pattern (step S203). After step S203, the control device 110 terminates the first priority execution process and ends the combination selection control without starting other priority controls (steps S300 to S700).

[0092] Furthermore, if the first priority is not combination selection A (NO in step S201), the control device 110 determines whether the first priority in the combination selection order is combination selection B (step S211). If the first priority is not combination selection B, the control device 110 proceeds to step S221. On the other hand, if the first priority is combination selection B, the control device 110 determines whether there is a weighing hopper 70 that is subject to zero correction and it is time to perform zero correction (step S212).

[0093] If there is no weighing hopper 70 to be zero-corrected, the control device 110 terminates the first priority execution process and then starts the second priority execution process (step S300). On the other hand, if there is a weighing hopper 70 to be zero-corrected, the control device 110 executes combination selection B. That is, the control device 110 sets the combination pattern that includes the weighing hopper to be zero-corrected and has the highest priority among the N combination patterns stored in area MN as the final combination pattern (step S213). After step S213, the control device 110 terminates the first priority execution process and ends the combination selection control without starting any other priority controls.

[0094] Furthermore, if the first priority is not combination selection B (NO in step S211), the control device 110 determines whether the first priority in the combination selection order (first priority) is combination selection C (step S221). If the first priority is not combination selection C, the control device 110 proceeds to step S231. On the other hand, if the first priority is combination selection C, the control device 110 determines whether any of the N combination patterns stored in area MN include a weighing hopper 70 and a memory hopper 80 of the same head (step S222).

[0095] If none of the N combination patterns stored in area MN contain the same head weighing hopper 70 and memory hopper 80, the control device 110 terminates the first priority execution process and then starts the second priority execution process (step S300). On the other hand, if any of the N combination patterns stored in area MN contain the same head weighing hopper 70 and memory hopper 80, the control device 110 executes combination selection C. That is, the control device 110 selects the highest-ranking combination pattern from among the N combination patterns stored in area MN that does not contain the same head weighing hopper 70 and memory hopper 80 as the final combination pattern (step S223). After step S223, the control device 110 terminates the first priority execution process and ends the combination selection control without starting any other priority controls.

[0096] Furthermore, if the first priority is not combination selection C (NO in step S221), the control device 110 determines whether the first priority in the combination selection order (first priority) is combination selection D (step S231). If the first priority is not combination selection D, the control device 110 proceeds to step S241. On the other hand, if the first priority is combination selection D, the control device 110 executes combination selection D. That is, the control device 110 selects the combination pattern with the fewest number of hoppers among the N combination patterns stored in area MN, and the highest priority among them (step S233). After step S233, the control device 110 terminates the first priority execution process and ends the combination selection control without starting any other priority controls.

[0097] Furthermore, if the first priority is not combination selection D (NO in step S231), the control device 110 determines whether the combination scale 1 is a semi-automatic combination scale and whether the first priority in the priority order for combination selection is combination selection E (step S241).

[0098] If the combination scale 1 is not a semi-automatic combination scale, or if the first priority is not combination selection E, the control device 110 proceeds to step S251. On the other hand, if the combination scale 1 is a semi-automatic combination scale and the first priority is combination selection E, the control device 110 executes combination selection E. That is, the control device 110 selects the combination pattern that does not include both weighing tanks of the weighing hopper 70 and has the highest priority among them as the final combination pattern (step S243). After step S243, the control device 110 terminates the first priority execution process and ends the combination selection control without starting any other priority controls.

[0099] Furthermore, if the first priority is not combination selection E (NO in step S241), the control device 110 determines whether the combination scale 1 is a semi-automatic combination scale and whether the first priority in the priority order for combination selection is combination selection F (step S251).

[0100] If the combination scale 1 is not a semi-automatic combination scale, or if the first priority is not combination selection F, the control device 110 terminates the first priority execution process and then starts the second priority execution process (step S300). On the other hand, if the combination scale 1 is a semi-automatic combination scale and the first priority is combination selection F, the control device 110 executes combination selection F. That is, the control device 110 prioritizes selecting a combination pattern that includes the hopper of the head near the center of the equipment where the loader stands as the final combination pattern (step S253). After step S253, the control device 110 terminates the first priority execution process and ends the combination selection control without starting any other priority controls.

[0101] Note that steps S241, S243, S251, and S253 can be omitted if the combination scale 1 is an automatic combination scale. This reduces the control load.

[0102] The following describes the second priority execution process performed by the control device 110, using the flowchart in Figure 7.

[0103] As shown in Figure 7, the content of the second priority execution process is generally the same as that of the first priority execution process. In Figure 7, steps in the second priority execution process that are the same as those in the first priority execution process are given the same step number, and their explanation is omitted.

[0104] The difference between the content of the second priority execution process and the content of the first priority execution process is that in steps S301, S311, S321, S331, S341, and S351, which correspond to steps S201, S211, S221, S231, S241, and S251 of the first priority execution process, the second priority is determined instead of the first priority.

[0105] Although not shown in the diagram, the contents of the third-priority execution process to the sixth-priority execution process are generally the same as those of the first-priority execution process. In other words, the difference between the contents of the third-priority execution process to the sixth-priority execution process and the first-priority execution process is that in each step corresponding to steps S201, S211, S221, S231, S241, and S251 of the first-priority execution process, the third-priority to sixth-priority processes are determined instead of the first-priority process.

[0106] Thus, in the main control shown in Figure 3, by executing the combination recording control and combination selection control shown in Figures 4 to 7, the combination scale 1 can acquire data for predetermined periods or for predetermined number of operations when combination selections A to F are executed for various conditions. The various conditions include, for example, the model of the combination scale, the type of object to be weighed, the properties of the object to be weighed, the target combination weight value, the selected hopper, the re-selected combination weight, the re-selected combination, the best combination weight, the weight values ​​of each hopper, and operating parameters.

[0107] The control device 110 can calculate the combined discharge weight value and operating data for the final combination pattern discharged from the hopper to the collection unit 90 by aggregating this data. The control device 110 can also calculate the combination accuracy, for example, by the average and variance values ​​for each predetermined period or each predetermined number of operations. Furthermore, the control device 110 can calculate the operating speed, for example, by the number of operating data obtained for each predetermined period or each predetermined number of operations.

[0108] In this way, the control device 110 can generate a learning model by learning the relationship between data (and furthermore, results obtained by aggregating data) at predetermined intervals or at predetermined operation cycles when combination selections A to F are performed for various conditions. By using such a learning model, the control device 110 can select combination selections A to F that improve combination accuracy based on various conditions that can be obtained in advance in the combination calculation. In addition, the control device 110 can select combination selections A to F that can achieve not only combination accuracy but also other effects (for example, improvement of operating speed, improvement of yield, etc.).

[0109] Next, the utilization phase will be explained. In the utilization phase, the control device 110 uses the learned model to operate the combination scale 1.

[0110] Specifically, the control device 110 performs the main control in the utilization phase shown in Figure 8. The main control in the utilization phase differs from the main control in the learning phase shown in Figure 3 in that steps S3 and S4 are omitted, and step S90 is performed instead of step S9.

[0111] In step S90, the control device 110 performs combination selection control using a learning model. That is, by using the learning model, the control device 110 selects combinations A to F that improve combination accuracy based on various conditions that can be obtained in advance in the combination calculation. Next, the control device 110 uses the selected combination to perform hopper control (step S11), etc., so that a suitable combination discharge weight value is discharged from the predetermined hoppers (weighing hopper 70 and memory hopper 80) to the collection unit 90.

[0112] As described above, in the combination scale 1 according to this embodiment, Multiple hoppers (weighing hopper 70 and memory hopper 80) for containing the supplied weighed material, A control device 110 that performs combined calculations based on the weight of the objects to be weighed, which are contained in a plurality of hoppers (weighing hopper 70 and memory hopper 80), A combination scale comprising, The control device 110 is In the aforementioned combination operation, Multiple (e.g., N) combination patterns are obtained such that the combined weight of the objects to be weighed falls within a predetermined weight range. From among the multiple aforementioned combination patterns, one combination pattern is selected based on the first condition (any of combination selections A to F).

[0113] This configuration can improve combination accuracy.

[0114] Furthermore, in the combination scale 1 according to this embodiment, The aforementioned combination patterns are prioritized for selection by comparing the combined weight of each object to be weighed with the target weight.

[0115] This configuration allows for further improvement in combination accuracy.

[0116] Furthermore, in the combination scale 1 according to this embodiment, The aforementioned first condition (any of combination selections A to F) is selected from among multiple first conditions (combination selections A to F) based on the second condition (learning model).

[0117] This configuration allows for further improvement in combination accuracy.

[0118] Furthermore, in the combination scale 1 according to this embodiment, The second condition (learning model) is set based on the combined weights of each combination pattern based on the multiple first conditions (combination selections A to F) (see, for example, Figures 6 and 7).

[0119] This configuration allows for further improvement in combination accuracy.

[0120] Furthermore, in the combination scale 1 according to this embodiment, The aforementioned second condition is comprised of a learning model that has learned the combined weights for each combination pattern based on the aforementioned multiple first conditions (combination selections A to F).

[0121] This configuration allows for further improvement in combination accuracy.

[0122] Although embodiments of the present invention have been described above, the present invention is not limited to the above configuration, and various modifications are possible within the scope of the invention as described in the claims.

[0123] For example, in this embodiment, six combination selections (combination selections A to F) are given as the first condition, but the number of combination selections is not limited to these. Furthermore, the content of the combination selections is not limited to those in this embodiment.

[0124] Furthermore, although the combination scale 1 according to this embodiment has a memory hopper 80, it does not necessarily have to have a memory hopper 80.

[0125] Furthermore, the training data used to generate the learning model is not limited to that according to this embodiment, and various types of data can be used. [Explanation of Symbols]

[0126] 1 Combination scale 70 Measuring hoppers 80 memory hopper 110 Control device

Claims

1. Multiple hoppers for containing the material to be weighed, A control device that performs a combination calculation based on the weight of the objects to be weighed, which are contained in a plurality of hoppers, A combination scale comprising, The control device is In the aforementioned combination operation, Multiple combination patterns are obtained such that the combined weight of the objects to be weighed falls within a predetermined weight range. From among the multiple aforementioned combination patterns, one combination pattern is selected based on the first condition. Combination scale.

2. The aforementioned combination patterns are selected based on a comparison between the combined weight of each object to be weighed and the target weight, thereby determining the priority of selection. The combination scale according to claim 1.

3. The aforementioned first condition is selected from among several first conditions based on the second condition. The combination scale according to claim 1.

4. The second condition is set based on the combined weights of each combination pattern based on the plurality of first conditions. The combination scale according to claim 1.

5. The second condition is comprised of a learning model that has learned the combined weights for each combination pattern based on the plurality of first conditions. The combination scale according to claim 4.

Images