Weighing system

Abstract

The present invention provides a weighing system capable of performing combinatorial calculations in an appropriate manner. [Solution] The system comprises a plurality of weighing hoppers 150 arranged along the circumference in a plan view and capable of weighing supplied articles; a combination calculation unit capable of performing article combination calculations using the weighed values ​​of the articles weighed in the plurality of weighing hoppers 150; a distributed feeder 120, a linear feeder 130, and a supply device 200 for transporting articles along a transport path from the article supply source to the weighing hoppers 150; and a control device 300 capable of controlling the operation of the distributed feeder 120, the linear feeder 130, and the supply device 200 based on image data of the articles supplied to the transport path.

Description

Technical Field

[0001] The present invention relates to the technology of weighing systems.

Background Art

[0002] Conventionally, there is known a combination scale that conveys an object to be weighed to a plurality of weighing units and performs a combination calculation of the object to be weighed using the measured weight values at each weighing unit. For example, it is as described in Patent Document 1.

[0003] Patent Document 1 describes a combination weighing device that measures the weight of an article (object to be weighed) conveyed on a conveyance path using a plurality of weighing hoppers (weighing units). In the combination weighing device, an article conveyed from a supply source of the article is input by a conveyance conveyor. The combination weighing device disperses the input article along a conical dispersion feeder and conveys the article conveyed from the dispersion feeder to a plurality of pool hoppers by a radial feeder extending radially from the dispersion feeder. The article conveyed from the radial feeder is temporarily stored in a plurality of pool hoppers and then discharged to a weighing hopper located below each pool hopper. Further, the combination weighing device can obtain a measured value of an article for each of the plurality of weighing hoppers, calculate a combined weight obtained by appropriately combining each measured value, and execute a combination calculation for selecting a combination of weighing hoppers in which the combined weight is within a predetermined weight range.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] In the above-described combination weighing device (combination scale), for example, it is desired to perform combination calculation preferably in order to improve combination accuracy and combination speed.

[0006] In conventional combination scales, the operating parameters of the conveying equipment, such as conveyor belts, distributed feeders, and radial feeders, which transport goods along the transport path, were pre-set. However, if, for example, there is variation in the amount of goods loaded on the radial feeder, the pre-set operating parameters may not be sufficient to cope, and it is anticipated that the amount of goods transported along the transport path will become unstable.

[0007] It is also conceivable that the operator could set the above operating parameters according to the amount of goods loaded on the radial feeder. However, since this method requires skilled operators, it is anticipated that it will not be able to adequately handle variations in the amount of goods on the radial feeder, and consequently, the amount of goods fed into the pool hopper may become unstable. In the above combination scale, if the amount of goods transported along the transport path (the amount of goods fed into the pool hopper) becomes unstable as described above, it may become difficult to perform combination calculations appropriately.

[0008] The present invention aims to provide a weighing system that can stabilize the amount of goods transported along a transport path and perform combination calculations appropriately, in order to solve the above problems. [Means for solving the problem]

[0009] To achieve the above objective, a weighing system according to one aspect of the present invention comprises: a plurality of weighing units arranged along the circumference in a plan view and capable of weighing a supplied object to be weighed; a combination calculation unit capable of performing combination calculations of the object to be weighed using the weighed values ​​of the object to be weighed measured by the plurality of weighing units; a transport device for transporting the object to be weighed along a transport path from the supply source of the object to be weighed to the weighing units; and a control device capable of controlling the operation of the transport device based on imaging results of the object to be weighed along the transport path. [Effects of the Invention]

[0010] According to the present invention, combinatorial calculations can be performed efficiently. [Brief explanation of the drawing]

[0011] [Figure 1] A schematic front view showing a weighing system according to one embodiment of the present invention. [Figure 2] A plan view showing the distributed feeder and linear feeder of a combination scale. [Figure 3] A block diagram showing the weighing system. [Figure 4] A schematic front view showing how goods are supplied to the linear feeder of a combination scale. [Figure 5] A flowchart illustrating the processes performed by the weighing system. [Figure 6] (a) A plan view showing an example of image learning data obtained by imaging objects on a physical simulation device, where the objects appear relatively few in number. (b) A plan view showing an example of image learning data where the objects appear relatively many in number. [Figure 7] A flowchart illustrating the model metricing process. [Figure 8] A flowchart illustrating the transport control process. [Figure 9] (a) A schematic diagram showing a straight feeder when the quantity of goods in section A is the standard quantity. (b) A schematic diagram showing a straight feeder when the quantity of goods in section A is below the lower limit of the standard quantity, and the quantity of goods in section B is the standard quantity. (c) A schematic diagram showing a straight feeder when the quantity of goods X in both section A and section B is below the lower limit of the standard quantity, but the total quantity of goods is above the lower limit of the standard quantity. [Figure 10] (a) A schematic diagram showing a straight feeder when the total quantity of goods is less than the lower limit of the standard quantity. (b) A schematic diagram showing a straight feeder when the quantity of goods in section A is less than the lower limit of the standard quantity, while the quantity of goods in section B is greater than the upper limit of the standard quantity. (c) A schematic diagram showing a straight feeder when the quantity of goods in section A is excess quantity Z, while the quantity of goods in section B is less than the lower limit of the standard quantity. [Figure 11] A flowchart illustrating the supply control process.

Best Mode for Carrying Out the Invention

[0012] Hereinafter, the front-rear direction, left-right direction, and up-down direction are defined according to the arrows shown in the drawings, respectively. Also, in each drawing, for convenience of explanation, the shape and size of the members may be appropriately changed.

[0013] First, the configuration of the weighing system 1 according to the first embodiment of the present invention will be described. The weighing system 1 weighs the weighed object from a predetermined supply source and discharges the weighed object by a predetermined weight. The weighing system 1 includes a combination scale 100, a supply device 200, a control device 300, an imaging device 400, a learning model generation unit 500, and a learning model execution unit 600.

[0014] The combination scale 100 shown in FIGS. 1 to 4 weighs the weighed object supplied from the supply device 200 described later by a predetermined amount. As the weighed object, various granular articles (for example, foods such as grains and dried fruits) can be adopted. Hereinafter, the weighed object may be simply referred to as "article" for explanation.

[0015] The combination scale 100 can be used in a packaging line or the like that puts the weighed articles into a packaging machine 170 (see FIG. 1) and packs them in bags. The combination scale 100 can weigh various types of articles. The combination scale 100 includes a base 110, a dispersion feeder 120, a straight feeder 130, a supply hopper 140, a weighing hopper 150, and an aggregating unit 160. Note that in FIGS. 1 and 4, each part of the combination scale 100 is schematically illustrated.

[0016] The base 110 constitutes the central part in the plan view of the combination scale 100. The base 110 is formed in a hollow shape capable of accommodating various devices inside. Note that in FIG. 1, the base 110 is schematically illustrated.

[0017] The dispersion feeder 120 disperses the articles supplied to the combination weigher 100 in the radial direction (horizontal direction) of the base body 110. The dispersion feeder 120 includes a top cone 121 having a substantially conical shape (or a frustum of a cone shape), and a vibration mechanism 122 that vibrates the top cone 121. The top cone 121 is provided above the base body 110 via the vibration mechanism 122 such that the top (upper end) is located at the central portion of the base body 110 in a plan view.

[0018] By vibrating the top cone 121 due to the operation of the vibration mechanism 122, the articles supplied to the dispersion feeder 120 are dispersed in the radial direction along the slope of the top cone 121 and fall downward (toward the straight feeder 130 described later) from the outer radial portion of the top cone 121. The dispersion feeder 120 according to the present embodiment can change the amplitude intensity (vibration amplitude) of the vibration mechanism 122 and the operating time (vibration time). In the present embodiment, by changing the parameters (operation parameters) of the amplitude intensity and the vibration time, the mode of dispersion of the articles by the dispersion feeder 120 can be changed.

[0019] The articles from the dispersion feeder 120 are supplied to a plurality of weighing units constituted by a straight feeder 130, a supply hopper 140, and a weighing hopper 150 described below. The weighing unit can weigh the articles dispersed by the dispersion feeder 120 by a predetermined amount. In the present embodiment, 14 weighing units are arranged along the circumferential direction of the base body 110.

[0020] The straight feeder 130 shown in FIGS. 1 to 4 conveys the articles dispersed by the dispersion feeder 120 so as to move straight outward in the radial direction of the base body 110. A plurality (14) of straight feeders 130 are arranged along the circumferential direction of the base body 110. The straight feeder 130 includes a straight trough 131 extending along the radial direction of the base body 110, and a vibration mechanism 132 that vibrates each straight trough 131. In FIG. 2, one straight feeder 130 located on the right side among the plurality of straight feeders 130 is shown filled in. Further, in FIG. 4, only one (right side) straight feeder 130 is illustrated.

[0021] Each straight trough 131 is formed in a trough shape, having a bottom and a pair of side walls rising from the bottom. Each straight trough 131 is arranged radially so as to be located radially outward of the distributed feeder 120 via each vibration mechanism 132. The straight trough 131 is formed in a shape according to the type of article.

[0022] The end of the straight trough 131 on the distributed feeder 120 side (hereinafter referred to as the "base end") is below the top cone 121 and is positioned to overlap with the radially outer part of the top cone 121 in a plan view (see Figure 4). The end of the straight trough 131 opposite to the base end (hereinafter referred to as the "tip") is located radially outward from the base body 110.

[0023] The vibration of the excitation mechanism 132 causes the straight trough 131 to vibrate, causing the articles supplied to the straight feeder 130 to move along the straight trough 131 radially outward from the base 110. The articles then fall from the tip of the straight trough 131 and are supplied to the supply hopper 140, which will be described later. In this embodiment, the amplitude intensity (vibration amplitude) and the operating time (vibration time) of the vibration excitation mechanism 132 of the straight feeder 130 can be changed. In this embodiment, the manner in which articles are transported by the straight feeder 130 can be changed by changing the parameters (operating parameters) of the amplitude intensity and vibration time.

[0024] The supply hopper 140 shown in Figures 1 and 4 temporarily stores the goods transported by the linear feeder 130 and then supplies the stored goods to the weighing hopper 150, which will be described later. Although omitted in Figures 1 and 4, multiple supply hoppers 140 are arranged along the circumferential direction on the radially outer side of the base body 110, corresponding to each of the multiple linear feeders 130 (14 in total). The supply hoppers 140 are supported by the base body 110 via appropriate support members.

[0025] The supply hopper 140 is formed in a roughly cylindrical shape with an opening facing upwards. The supply hopper 140 is designed so that its bottom can be opened and closed by an appropriate drive mechanism (not shown). When the bottom of the supply hopper 140 is opened, the items inside the supply hopper 140 fall out and are supplied to the weighing hopper 150.

[0026] The weighing hopper 150 shown in Figure 1 measures the weight of items supplied from the supply hopper 140 and supplies the weighed items to the collection unit 160, which will be described later. The weight of the items in the weighing hopper 150 is measured using a weight sensor 151. A suitable load cell can be used as the weight sensor 151. Although omitted in Figure 1, multiple weighing hoppers 150 (14 in total) are arranged along the circumferential direction on the radially outer side of the base body 110, so as to be located below each of the multiple supply hoppers 140. The weighing hoppers 150 are supported by the base body 110 via suitable support members.

[0027] The weighing hopper 150 is formed in a roughly cylindrical shape with an opening facing upwards. The weighing hopper 150 is configured so that its bottom can be opened and closed by an appropriate drive mechanism (not shown). When the bottom of the weighing hopper 150 is opened, the items in the weighing hopper 150 fall out and are supplied to the collection unit 160.

[0028] The collection unit 160 shown in Figure 4 collects articles supplied from the weighing hopper 150 and is capable of loading the articles into the packaging machine 170. The collection unit 160 has a collection hopper (not shown) for temporarily storing articles from the weighing hopper 150. The collection unit 160 discharges the stored articles to the packaging machine 170 based on a discharge command from, for example, the packaging machine 170.

[0029] The configuration of the combination scale 100 has been described above. The configuration of the combination scale 100 is not limited to the example described above, and it is possible to add appropriate sensors or other equipment. For example, the combination scale 100 may be equipped with an item detection sensor (not shown) capable of detecting the level (height) of items in the distributed feeder 120 or the linear feeder 130.

[0030] The supply device 200 shown in Figure 1 is capable of supplying goods from a predetermined supply source to a combination scale 100. The supply device 200 comprises a bucket lifter (not shown) for transporting goods to a predetermined height position, and a supply feeder 210 for transporting goods from the bucket lifter and feeding them into the combination scale 100. The supply device 200 is configured to allow the position of the supply feeder 210 relative to the distribution feeder 120 of the combination scale 100 (the position where goods are fed into the distribution feeder 120) to be changed. By changing the position where goods are fed into the distribution feeder 120, it becomes easier to supply goods to any linear feeder 130. The supply feeder 210 is equipped with a vibration mechanism (not shown) for vibrating the supply feeder 210. In this embodiment, the supply device 200 allows setting the amplitude intensity (vibration amplitude) and the operating time (vibration time). The supply device 200 also allows changing the operating speed of the bucket lifter. In this embodiment, the manner in which the supply device 200 supplies articles can be changed by changing the above-mentioned parameters (operating parameters) of amplitude intensity, vibration time, and operating speed.

[0031] The control device 300 shown in Figure 3 performs various calculations related to the operation of the combination scale 100 and the supply device 200. The control device 300 is equipped with a processing unit such as a CPU and a memory device. The control device 300 can control the operation of each mechanism of the combination scale 100 (such as the vibration mechanism 122 and vibration mechanism 132, the drive mechanism of the supply hopper 140 and weighing hopper 150, etc.) and perform calculations based on the measured values ​​of the weight sensor 151. The control device 300 can also control the operation of the bucket lifter and the supply feeder 21 of the supply device 200. Furthermore, the control device 300 can set the operating parameters of the combination scale 100 (dispersion feeder 120 and linear feeder 130) and the supply device 200 by operating a predetermined operating device (not shown).

[0032] The combination scale 100, the supply device 200, and the control device 300 have been described above. The control device 300 can supply and weigh goods by controlling the combination scale 100 and the supply device 200. The control modes of the supply device 200 and the control device 300 will be described below using Figure 1. The control mode described below will be the one without using the learning model described later.

[0033] The control device 300 can supply a preset amount of goods to the combination scale 100 by controlling the supply device 200. More specifically, the control device 300 can load any amount of goods into any position on the distribution feeder 120 by changing the position of the supply feeder 210 and controlling the transport of goods by the supply feeder 210.

[0034] Furthermore, the control device 300 can weigh the items fed in by the supply device 200 and control the discharge of a predetermined weight of items to the packaging machine 170. More specifically, the control device 300 first operates the distributed feeder 120 to transport the items fed into the combination scale 100 to each weighing unit (straight feeder 130, supply hopper 140, and weighing hopper 150).

[0035] Next, the control device 300 operates the linear feeders 130 of each weighing unit to transport the items on the linear troughs 131 to the supply hopper 140. As the items on the linear troughs 131 are transported, the items loaded on the distributed feeders 120 move onto the linear troughs 131. Next, the control device 300 loads items from the supply hopper 140 into the weighing hopper 150 and weighs the items in each weighing hopper 150.

[0036] Next, the control device 300 acquires the weighing value of each item (measured value based on the weight sensor 151) for each weighing unit, calculates the total weight (combined weight) by appropriately combining each weighing value, and can perform control (combination calculation) to select a combination of weighing units (weighing hoppers 150) in which the combined weight falls within a predetermined weight range. The control device 300 also discharges the items from the weighing units (weighing hoppers 150) that are to be discharged as a result of the combination calculation to the packaging machine 170. In this way, the combination scale 100 can discharge a predetermined weight of items to the packaging machine 170.

[0037] During the above combination calculation process, the supply hopper 140, into which the goods were loaded for the weighing hopper 150, becomes empty. In this case, the control device 300 supplies goods via the supply device 200 and operates the distributed feeder 120 and the linear feeder 130 of the target weighing unit to transport the goods to the target supply hopper 140. By executing the above-described control, the control device 300 can perform combination calculations continuously.

[0038] In the combination scale 100 described above, it is desirable to perform combination calculations appropriately in order to improve, for example, combination accuracy and combination speed. To improve combination accuracy and combination speed in combination calculations, it is desirable to control the weight of the items fed from the linear feeder 130 to the supply hopper 140 so that it becomes the optimal weight. Specifically, it is desirable to control the weight of the items fed from the linear feeder 130 to the supply hopper 140 so that the weight of the items in the weighing hopper 150 participating in the combination calculation becomes the optimal weight (the weight that results in the most combinations).

[0039] In typical combination calculations, the operating parameters of the supply device 200, the distributed feeder 120, and the linear feeder 130 were pre-set. However, if, for example, the type of goods supplied to the combination scale 100 or the amount of goods supplied from the supply source to the supply device 200 changes, variations may occur in the amount of goods loaded onto the linear feeder 130. If such variations in the amount of goods occur, the pre-set operating parameters may not be sufficient to cope, and it is anticipated that the amount of goods fed into the supply hopper 140 will become unstable.

[0040] It is also conceivable that the operator could set the above operating parameters according to the amount of goods loaded on the linear feeder 130. However, since this method requires skilled operators, it is anticipated that it will not be able to adequately handle variations in the amount of goods on the linear feeder 130, and consequently, the amount of goods fed into the supply hopper 140 may become unstable.

[0041] Furthermore, in the combination scale 100, the quantity (or form) of items loaded on the linear feeder 130 (linear trough 131) changes moment by moment. Therefore, if the linear feeder 130 is operated with a single preset parameter, it is anticipated that it will be difficult to stabilize the quantity of items fed into the supply hopper 140.

[0042] Here, instead of the operator setting the operating parameters of the supply device 200, it is also conceivable to implement control that automatically switches the operation of the supply device 200 (supply feeder 210) ON and OFF based on, for example, the detection result of the item detection sensor (detection result of the level of items on the distributed feeder 120 or the linear feeder 130). However, in the above method, since the operation of the supply feeder 210 is switched ON and OFF intermittently, there is a possibility that variations may occur in the items loaded on the linear feeder 130, and consequently, variations may occur in the weight of the items fed into the supply hopper 140.

[0043] The following describes an example of how the items loaded on the straight feeder 130 (straight trough 131) may be scattered. In the following description, as shown in Figure 4, the straight trough 131 will be divided into two parts, and the arrangement of items on the straight trough 131 will be described accordingly. Specifically, the front half of the straight trough 131 will be referred to as "Section A," and the base half of the straight trough 131 will be referred to as "Section B."

[0044] When the total amount of goods loaded on the straight trough 131 (sections A and B) is relatively small, or when the amount of goods loaded on section A of the straight trough 131 is relatively small, the amount of goods fed into the supply hopper 140 tends to be small. On the other hand, when the total amount of goods loaded on the straight trough 131 (sections A and B) is relatively large, or when the amount of goods loaded on section A of the straight trough 131 is relatively large, the amount of goods fed into the supply hopper 140 tends to be large. Thus, it is expected that the amount of goods fed into the supply hopper 140 will vary depending on the type of goods loaded on the straight trough 131.

[0045] Here, it is conceivable to suppress variations in the amount of goods fed into the supply hopper 140 by setting the operating parameters of the linear feeder 130 according to the weight of the goods in sections A and B of the linear feeder 130. Specifically, it is conceivable to stabilize the amount of goods fed into the supply hopper 140 by installing a weight sensor capable of weighing the goods on each of the linear troughs 131 and setting the operating parameters of the linear feeder 130 based on the detected values ​​of the weight sensors. However, in this case, it is expected that the cost of installing multiple weight sensors will increase. In addition, it is difficult to accurately weigh the goods in each section of a single linear trough 131.

[0046] In the weighing system 1 according to this embodiment, the amount of goods fed into the supply hopper 140 can be stabilized by controlling the supply device 200 and the straight feeder 130 based on the estimation results of a learning model that can estimate the amount of goods loaded on the straight trough 131 (section A and section B) using image data. Below, the imaging device 400, the learning model generation unit 500, and the learning model execution unit 600 used for generating and executing the learning model will be described.

[0047] The imaging device 400 is capable of imaging items on the distribution feeder 120 and linear feeder 130 of the combination scale 100. The imaging device 400 is positioned to image the distribution feeder 120 and linear feeder 130 from above. Various cameras can be used for the imaging device 400. By imaging with the imaging device 400, image data, which is the result of imaging items loaded on the linear trough 131 of the linear feeder 130, can be obtained. Imaging by the imaging device 400 is performed under the control of the control device 300.

[0048] The learning model generation unit 500 is capable of generating a learning model. Here, a learning model is a model that can output an estimated quantity (a physical quantity equivalent to weight, described later) indicating the quantity of items on the linear feeder 130 based on image data of the linear feeder 130. The learning model generation unit 500 can be an information processing device (for example, a personal computer) capable of processing various types of information. The learning model generation unit 500 stores a program for generating a learning model and a program for physical simulation (physics calculation engine), described later.

[0049] The learning model is generated by machine learning. Here, machine learning refers to the process of analyzing training data (weight training data and image training data described later), learning the patterns of features extracted from the training data, and creating a model that enables appropriate processing. As a machine learning method, technologies using AI (artificial intelligence) can be employed. In this embodiment, as a machine learning method, 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, is employed. However, the machine learning method is not limited to the example described above, and various methods such as decision trees and random forests can be employed.

[0050] The learning model execution unit 600 executes the learning model generated by the learning model generation unit 500. The learning model execution unit 600 is equipped with computing devices such as a CPU and GPU, and a memory device. By executing the learning model in the learning model execution unit 600, it is possible to estimate the quantity of goods loaded on the straight trough 131. The explanation of how to estimate the quantity of goods using the learning model will be described later.

[0051] The configuration of weighing system 1 has been described above. Below, we will explain each process performed by weighing system 1 using Figures 5 to 11.

[0052] The flowchart in Figure 5 shows an overview of the processes performed by the weighing system 1. As shown in step S10 of Figure 5, the learning model generation unit 500 of the weighing system 1 acquires learning data for items to generate a learning model. Here, in generating the learning model, information on a specific type of item is used as the item to be weighed. In the following, the item used to generate the learning model will be referred to as "item X".

[0053] The above training data includes "weight training data" and "image training data" for item X. The "weight training data" is data (weight simulation results) corresponding to the weight of item X loaded on the straight trough 131 of the straight feeder 130. The weight training data includes data on the weight of item X in sections A and B of the straight trough 131. The "image training data" is data corresponding to the imaging results of item X loaded on the straight trough 131 of the straight feeder 130.

[0054] In this embodiment, the learning model generation unit 500 acquires weight learning data and image learning data of item X through physical simulation. Here, physical simulation is a technique for creating images, etc., by simulating physical phenomena. In this embodiment, instead of using the weight and imaging results of item X loaded on the straight feeder 130 (straight trough 131) of the actual combination scale 100 as learning data, the learning model is generated using weight learning data and image learning data obtained as a result of a physical simulation that mimics the combination scale 100 (dispersion feeder 120 and straight feeder 130).

[0055] The learning model generation unit 500 receives various information necessary for physical simulation (information on the shape and operation of each part of the combination scale 100 and the supply device 200, and information on the shape and weight of item X, etc.). Based on the above input information, the learning model generation unit 500 performs a simulation of combination calculations using the combination scale 100 and acquires the image capture results of item X on the straight feeder 130 (straight trough 131) at a certain point in time as image learning data.

[0056] Figure 6 shows an example of image learning data. The image learning data is generated as images of items X on the distributed feeder 120 and the linear feeder 130, captured from above, corresponding to the imaging results by the imaging device 400. Figure 6(a) shows image learning data when there are relatively few items X on the distributed feeder 120 and the linear feeder 130. Figure 6(b) shows image learning data when there are relatively many items X on the distributed feeder 120 and the linear feeder 130.

[0057] Furthermore, the learning model generation unit 500 acquires the simulation results of the weight of item X in sections A and B at the time image learning data is acquired as weight learning data. The learning model generation unit 500 acquires weight learning data and image learning data until the amount of data necessary for model generation is accumulated.

[0058] When the above training data is available, the learning model generation unit 500 generates a learning model, as shown in step S20 of Figure 5. The learning model generation unit 500 learns the relationship (correlation) between the feature quantities contained in the image training data of item X in the physical simulation and the weight training data. Specifically, when there are few images showing item X in the image training data (see Figure 6(a)), the weight training data for item X tends to show a relatively small weight. On the other hand, when there are many images showing item X in the image training data (see Figure 6(b)), the weight training data for item X tends to show a relatively large weight. The learning model generation unit 500 generates a learning model by learning the correlation between the image training data and the weight training data as described above. The learning model generation unit 500 can update the weights of the neural network by backpropagating the error between the output result of the learning model based on the image training data (estimated value of the amount of item) and the training data, and advance the learning process.

[0059] The learning model according to this embodiment can output an estimated quantity, which is an estimated value of the amount of items (e.g., item X) on the linear feeder 130, based on image data of the linear feeder 130 actually captured by the imaging device 400. In this embodiment, the "estimated quantity" is not the weight of the items on the linear feeder 130 itself, but a quantity correlated with the weight of the items (a physical quantity equivalent to weight). As the above estimated quantity, it is possible to adopt a value that correlates with both image learning data and weight learning data, regardless of the shape or density of the items. With the above configuration, for example, when estimating the amount of item Y, which has a different shape and density from item X, the learning model generated based on item X can be used as is to estimate the amount of item Y. In this embodiment, an example is shown in which information of item X on the linear feeder 130 is used as the information to be learned when generating the learning model and as the information used when executing the learning model, but it is not limited to the above example, and information of item X on the distributed feeder 120 may also be used in the learning and execution of the learning model.

[0060] Once the above-mentioned learning model is generated, the weighing system 1 (control device 300) executes a "model weighing process." Here, the "model weighing process" is a process that uses the learning model to control the combination scale 100 and the supply device 200 for combination calculations.

[0061] As shown in step S30 of Figure 5, the control device 300 can perform model weighing on item X based on the learned model. As mentioned above, the learned model according to this embodiment is also applicable to item Y, which is of a different type from item X. Therefore, as shown in step S40, the control device 300 can also perform model weighing on item Y based on the learned model.

[0062] The following explanation uses Figures 7 through 11 to describe the model weighing process. The following example describes a combination calculation for item X. Furthermore, the following explanation shows an example where item X is loaded onto the linear feeder 130 and the model weighing process is performed. The model weighing process is executed at regular intervals (e.g., a few seconds or a few minutes).

[0063] In step S31, as shown in the flowchart of Figure 7, the control device 300 acquires image data on the combination scale 100 (see Figure 4). That is, the control device 300 uses the imaging device 400 to capture image data of the items X loaded on the straight trough 131 and acquires image data. After executing the process in step S31, the control device 300 proceeds to the process in step S32.

[0064] In step S32, the control device 300 determines whether or not there is a supply hopper 140 that needs to transport item X. That is, if there is a weighing unit whose supply hopper 140 has become empty due to the operation of the combination scale 100 during the execution of the combination calculation, the control device 300 determines that there is a supply hopper 140 that needs to transport item X. If the control device 300 determines that there is a supply hopper 140 that needs to transport item X, it proceeds to the process in step S33. On the other hand, if the control device 300 determines that there is no supply hopper 140 that needs to transport item X, it proceeds to the process in step S35.

[0065] In step S33, the control device 300 performs a "transport control process" on the linear feeder 130 that transports the item X to the empty supply hopper 140. Here, the "transport control process" is the process of controlling the linear feeder 130 based on the estimated quantity output by the learning model to feed the item X into the supply hopper 140. The transport control process will be explained below using the flowchart in Figure 8.

[0066] In the transport control process, the control device 300 controls the operation of the linear feeder 130 (linear trough 131) based on the relationship between the estimated amount of goods X loaded on the linear feeder 130 and a preset "target transport weight S". Here, "target transport weight S" is the target amount of goods X to be fed into the supply hopper 140. The target transport weight S is set as a physical quantity equivalent to weight. In addition, in the transport control process, "transport time T" and "transport intensity F" are preset as reference values ​​for the operation of the linear feeder 130.

[0067] Here, "transport time T" is the value set as the appropriate transport time when a "standard quantity" of goods is loaded into section A shown in Figure 4. Here, "standard quantity" is a value that represents a certain range of quantity based on the target transport weight S. In this embodiment, the standard quantity is set in the range of S / 2 or more and 3S / 2 or less. Note that the larger the transport time value, the more goods X can be transported.

[0068] Furthermore, "conveying strength F" is a value set as the appropriate conveying strength when a standard amount of goods X are loaded in section A. Note that the higher the conveying strength value, the faster goods X can be transported, and consequently, the more goods X can be transported. In this embodiment, when a standard amount of goods are loaded in section A, operating the straight feeder 130 at conveying strength F for a transport time T will supply an appropriate amount of goods to the supply hopper 140.

[0069] The following describes each process shown in the flowchart of Figure 8. In the transport control process, the control device 300 controls the learning model execution unit 600 to output an estimated value from the learning model based on the image data acquired in step S31 of Figure 7. The estimated value from the learning model is output for each of the multiple linear feeders 130. In addition, for each linear feeder 130, the estimated value from the learning model is output for section A, section B, and the overall value (sections A and B).

[0070] In step S101, the control device 300 determines whether the estimated quantity for section A is within the standard range (S / 2 or greater, and 3S / 2 or less). If the control device 300 determines that the estimated quantity for section A is within the standard range, it proceeds to step S102. On the other hand, if the control device 300 determines that the estimated quantity for section A is outside the standard range, it proceeds to step S103.

[0071] If the process proceeds to step S102, the amount of item X in section A is indicated to be the standard amount (see Figure 9(a)). In this case, the control device 300 operates the straight feeder 130 at a transport intensity F for the transport time T. After executing the process in step S102, the control device 300 terminates the transport control process.

[0072] In step S103, the control device 300 determines whether the estimated quantity for section A is less than the lower limit of the reference quantity (less than S / 2). If the control device 300 determines that the estimated quantity for section A is less than the lower limit of the reference quantity, it proceeds to the process in step S104. On the other hand, if the control device 300 determines that the estimated quantity for section A is equal to or greater than the lower limit of the reference quantity, it proceeds to the process in step S111.

[0073] In step S104, the control device 300 determines whether the estimated quantity for section B is within the standard range (S / 2 or greater, and 3S / 2 or less). If the control device 300 determines that the estimated quantity for section B is within the standard range, it proceeds to step S105. On the other hand, if the control device 300 determines that the estimated quantity for section B is outside the standard range, it proceeds to step S106.

[0074] If the process proceeds to step S105, it is indicated that the amount of item X in section A is less than the lower limit of the standard amount, and the amount of item X in section B is the standard amount (see Figure 9(b)). In this case, it is estimated that if the straight feeder 130 is operated with normal operating parameters (similar to when the standard amount of item X is loaded in section A), the amount of item fed into the supply hopper 140 will be less. Therefore, in step S105, the control device 300 operates the straight feeder 130 at a transport intensity of F for a transport time of 2T (twice the transport time T), or operates the straight feeder 130 at a transport intensity of 2F (twice the transport intensity F) for a transport time T. Note that it is possible to set either the transport time or the transport intensity to double as appropriate. After executing the process in step S105, the control device 300 terminates the transport control process.

[0075] In step S106, the control device 300 determines whether the estimated quantity for section B is less than the lower limit of the reference quantity (less than S / 2). If the control device 300 determines that the estimated quantity for section B is less than the lower limit of the reference quantity, it proceeds to step S107. On the other hand, if the control device 300 determines that the estimated quantity for section B is equal to or greater than the lower limit of the reference quantity, it proceeds to step S110.

[0076] In step S107, the control device 300 determines whether the estimated value of the entire straight trough 131 (sections A and B) is greater than or equal to the lower limit of the reference quantity (S / 2 or greater). If the control device 300 determines that the estimated value of the entire straight trough 131 is greater than or equal to the lower limit of the reference quantity, it proceeds to the process in step S108. On the other hand, if the control device 300 determines that the estimated value of the entire straight trough 131 is less than the lower limit of the reference quantity, it proceeds to the process in step S109.

[0077] If the process proceeds to step S108, the amount of goods X in section A and the amount of goods X in section B are both below the lower limit of the standard amount, but the total amount of goods X in the straight trough 131 is greater than or equal to the lower limit of the standard amount (see Figure 9(c)). In this case, because the total amount of goods X in the straight trough 131 is small, it is estimated that if the straight feeder 130 is operated with normal operating parameters, the amount of goods fed into the supply hopper 140 will be small. Therefore, in step S108, the control device 300 operates the straight feeder 130 at a transport intensity of 2F for a transport time of 2T. After executing the process in step S108, the control device 300 terminates the transport control process.

[0078] Furthermore, if the process proceeds to step S109, it is indicated that the total amount of goods X in the straight trough 131 is less than the lower limit of the reference amount (see Figure 10(a)). In this case, the control device 300 does not operate the straight feeder 130. After executing the process in step S109, the control device 300 terminates the transport control process.

[0079] Furthermore, if the process proceeds to step S110, it is shown that the amount of item X in section A is less than the lower limit of the standard amount, while the amount of item X in section B is greater than the upper limit of the standard amount (see Figure 10(b)). Hereinafter, the estimated amount for each section that is greater than the upper limit of the standard amount will be referred to as the "excess amount Z". In this case, it is estimated that if the straight feeder 130 is operated with normal operating parameters, the amount of items fed into the supply hopper 140 will be relatively small. Therefore, in step S110, the control device 300 operates the straight feeder 130 at a transport intensity F for a time obtained by multiplying the transport time T by 2S / Z, or operates the straight feeder 130 at an intensity obtained by multiplying the transport intensity F by 2S / Z for the transport time T. Note that it is possible to set as appropriate whether to multiply the transport time or the transport intensity by 2S / Z. After executing the process in step S110, the control device 300 terminates the transport control process.

[0080] Furthermore, if the process proceeds to step S111, it is shown that the amount of item X in section A is an excess amount Z, while the amount of item X in section B is below the lower limit of the standard amount (see Figure 10(c)). In this case, it is estimated that if the straight feeder 130 is operated with normal operating parameters, the amount of item fed into the supply hopper 140 will be relatively large. Therefore, in step S111, the control device 300 operates the straight feeder 130 at a transport intensity F for a time obtained by multiplying the transport time T by S / Z, or operates the straight feeder 130 at an intensity obtained by multiplying the transport intensity F by S / Z for the transport time T. Note that it is possible to set as appropriate whether to multiply the transport time or the transport intensity by S / Z. After executing the process in step S111, the control device 300 terminates the transport control process.

[0081] The processes of the transport control process have been described above. According to the above process, the operation of the linear feeder 130 can be controlled according to the quantity and distribution of items X loaded on the linear trough 131, thereby optimizing the amount of items fed into the supply hopper 140. This improves the weighing accuracy and weighing speed of the combination scale 100. In addition, in this embodiment, the operation of the linear feeder 130 is controlled according to the estimated quantity of a learning model based on image data. This eliminates the need to install a weight sensor on each linear feeder 130, unlike when the operation of the linear feeder 130 is controlled based on the actual weight of the items X on the linear trough 131, thus suppressing cost increases.

[0082] As shown in the flowchart of Figure 7, the control device 300 performs the transport control process shown in step S33, and then proceeds to the process in step S34.

[0083] In step S34, the control device 300 performs a "supply control process" on the straight feeders 130 that are not subject to the transport control process in step S33. Here, the "supply control process" is a process that controls the supply devices 200 and distributed feeders 120 based on the estimated values ​​of the learning model to supply goods X to the straight feeders 130 that are lacking goods X. In this embodiment, the control device 300 sets the operating parameters of the supply devices 200 and distributed feeders 120 based on the estimated values ​​of the learning model, thereby supplying goods X to sections A, B, and the entire straight trough 131 (sections A and B). Specifically, the control device 300 can supply goods X to sections A, B, and the entire straight trough 131 by outputting supply signals to the supply feeder 210 and the distribution feeder 120 that indicate operating parameters such as the position where goods X are fed into the distribution feeder 120 by the supply feeder 210 and the timing for operating the distribution feeder 120, so that goods X can be supplied to any position on the straight trough 131 (section A, section B, and the entire section).

[0084] The supply control process will be explained below using the flowchart in Figure 11. In the supply control process, the control device 300 executes each process based on the estimated values ​​of the learning model used in the transport control process in step S33. In the supply control process, the operation of the supply device 200 is controlled based on the relationship between the estimated values ​​of items X loaded on the straight feeder 130 (sections A and B of the straight trough 131) and a reference value based on the target transport weight S (a reference value similar to that used in the transport control process).

[0085] In step S201, the control device 300 determines whether the estimated quantity of section A of the straight trough 131 is less than the lower limit of the reference quantity (less than S / 2). If the control device 300 determines that the estimated quantity of section A is less than the lower limit of the reference quantity, it proceeds to the process in step S202. On the other hand, if the control device 300 determines that the estimated quantity of section A is equal to or greater than the lower limit of the reference quantity, it proceeds to the process in step S205.

[0086] In step S202, the control device 300 determines whether the estimated quantity of section B of the straight trough 131 is less than the lower limit of the reference quantity (less than S / 2). If the control device 300 determines that the estimated quantity of section B is less than the lower limit of the reference quantity, it proceeds to the process in step S203. On the other hand, if the control device 300 determines that the estimated quantity of section B is equal to or greater than the lower limit of the reference quantity, it proceeds to the process in step S204.

[0087] In step S205, the control device 300 determines whether the estimated quantity of section B of the straight trough 131 is less than the lower limit of the reference quantity (less than S / 2). If the control device 300 determines that the estimated quantity of section B is less than the lower limit of the reference quantity, it proceeds to the process in step S206. On the other hand, if the control device 300 determines that the estimated quantity of section B is equal to or greater than the lower limit of the reference quantity, it proceeds to the process in step S207.

[0088] If the process proceeds to step S203, it indicates that the quantity of item X in section A and the quantity of item X in section B are both below the lower limit of the reference quantity. In other words, if the process proceeds to step S203, it indicates that there is a shortage of item X in both section A and section B. In this case, the control device 300 outputs a supply signal to supply item X to both section A and section B, and operates the supply device 200 and the distributed feeder 120. After executing the process in step S203, the control device 300 terminates the supply control process.

[0089] If the process proceeds to step S204, it will be indicated that the amount of item X in section A is below the lower limit of the standard amount, while the amount of item X in section B is above the lower limit of the standard amount. In other words, if the process proceeds to step S204, it will be indicated that only section A is short of item X. In this case, the control device 300 outputs a supply signal to supply item X to section A and operates the supply device 200 and the distributed feeder 120. After executing the process in step S204, the control device 300 terminates the supply control process.

[0090] If the process proceeds to step S206, it indicates that the amount of item X in section A is above the lower limit of the standard quantity, while the amount of item X in section B is below the lower limit of the standard quantity. In other words, if the process proceeds to step S204, it indicates that only section B is short of item X. In this case, the control device 300 outputs a supply signal to supply item X to section B and operates the supply device 200 and the distributed feeder 120. After executing the process in step S206, the control device 300 terminates the supply control process.

[0091] If the process proceeds to step S207, it indicates that the quantity of item X in section A and the quantity of item X in section B are both above the lower limit of the reference quantity. In other words, if the process proceeds to step S207, it indicates that there is no shortage of item X in both section A and section B. In this case, the control device 300 does not output a supply signal to the supply device 200. After executing the process in step S207, the control device 300 terminates the supply control process.

[0092] The processes of the transport control process in step S34 shown in Figure 7 have been described above. According to the above process, if there is a shortage of goods X in the straight feeder 130 which is not subject to the transport control process, goods X can be supplied to the location where there is a shortage of goods X. This makes it possible to optimize the amount of goods fed into the supply hopper 140 and improve the weighing accuracy and weighing speed of the combination scale 100.

[0093] In the example described above, supply control processing was performed on the straight feeders 130 that were excluded from the transport control processing in step S33, but the system is not limited to the example described above. For example, in addition to the straight feeders 130 that were excluded as described above, supply control processing may also be performed on the straight feeders 130 that were subject to processing in step S109 of the transport control processing (straight feeders 130 in which the total amount of goods X in the straight trough 131 was small and transport control was not performed).

[0094] Furthermore, the example of supply control processing described above shows an example in which both the supply device 200 and the distributed feeder 120 are controlled to supply article X to an arbitrary position on the straight trough 131, but the method is not limited to the example described above. For example, either the supply device 200 or the distributed feeder 120 may be controlled to supply article X to an arbitrary position on the straight trough 131.

[0095] Furthermore, as shown in the flowchart of Figure 7, the control device 300 terminates the model metering process after executing the supply control process shown in step S34.

[0096] Furthermore, in step S35, which is the step to which the control device 300 proceeds if it is determined in step S32 that there are no supply hoppers 140 that need to transport goods, the control device 300 executes supply control processing for all linear feeders 130. Note that the content of the supply control processing in step S35 is the same as the supply control processing in step S34, except that the control is performed for all linear feeders 130, so the explanation is omitted. After executing the supply control processing shown in step S35, the control device 300 terminates the model weighing process.

[0097] According to the supply control process in step S35, if the transport control process is not executed and there is a shortage of goods X in the straight feeder 130, goods X can be supplied to the location where there is a shortage of goods X.

[0098] The control processes performed by the weighing system 1 have been described above. Note that the control described in this embodiment is an example, and the control performed by the weighing system 1 is not limited to the example described above; any additional or modified processes may be added. Furthermore, the specific numerical values ​​exemplified in the above description are examples and can be changed at will.

[0099] As described above, the weighing system 1 according to this embodiment is Multiple weighing hoppers 150 (weighing hoppers) are arranged along the circumference in a plan view and are capable of weighing supplied items (items to be weighed), A combination calculation unit (control device 300) capable of performing combination calculations of the items using the weighed values ​​of the items weighed in multiple weighing hoppers 150, In the transport path from the source of the article to the weighing hopper 150, the transport device (distributed feeder 120, linear feeder 130, and supply device 200) transports the article, A control device 300 capable of controlling the operation of the transport device (distributed feeder 120, linear feeder 130, and supply device 200) based on image data (imaging results) of the articles supplied to the transport path (on the linear feeder 130), It is equipped with the following features. This configuration allows for efficient combination calculations. Specifically, by controlling the operation of the conveying device based on image data of the items, the amount of items supplied to the weighing hopper 150 can be optimized from the viewpoint of efficiently performing combination calculations, such as improving the accuracy of the combination calculations.

[0100] Furthermore, the conveying device includes, A distribution feeder 120 that distributes the supplied articles horizontally, Multiple linear feeders 130 (conveyor feeders) transport the articles dispersed by the distribution feeder 120 to multiple weighing hoppers 150, It includes, The control device 300 is The operation of the conveying device (distributed feeder 120, linear feeder 130, and supply device 200) is controlled based on the estimated quantity of the items on the linear feeder 130, which is estimated based on the image data of the items loaded on the linear feeder 130. By configuring the device in this way, the operation of the conveying apparatus can be controlled based on the estimated quantity of items on the straight feeder 130 located upstream of the weighing hopper 150 in the conveying path, thereby optimizing the amount of items supplied to the weighing hopper 150.

[0101] Furthermore, the aforementioned estimate is This is output by a learning model generated by learning the relationship between the image learning data of the aforementioned article (information related to the imaging results) and the weight learning data of the aforementioned article (information related to the weight). By configuring it in this way, the operation of the transport device can be controlled based on the estimated values ​​from the learning model.

[0102] Furthermore, the weighing system 1 is, The system includes a learning model generation unit 500 (physical simulation unit) that acquires the image learning data and the weight learning data through physical simulation. By configuring the system in this way, image learning data and weight learning data can be easily obtained using physical simulations without actually acquiring image data or measuring the weight of objects. Furthermore, values ​​that are difficult to measure in reality, such as the weight learning data for sections A and B, can be obtained.

[0103] Furthermore, the conveying device includes, The aforementioned linear feeder 130 is included, The control device 300 is By controlling the operation of the linear feeder 130 based on the estimated amount, the amount of goods supplied from the linear feeder 130 to the supply hopper 140 (downstream of the transport path) can be adjusted. By configuring the system in this way, the operation of the linear feeder 130 located upstream of the weighing hopper 150 can be controlled to optimize the amount of goods supplied to the weighing hopper 150.

[0104] Furthermore, the learning model is The estimated quantities for section A and section B (multiple parts) that divide the linear feeder 130 in the conveying direction can be output, respectively. The control device 300 is The operation of the linear feeder 130 can be controlled based on the estimated values ​​for each of the aforementioned sections A and B. This configuration allows for efficient adjustment of the amount of goods supplied to the weighing hopper 150.

[0105] Furthermore, the conveying device includes, The aforementioned distributed feeder 120, A supply device 200 that supplies the articles from the supply source to the distributed feeder 120, It includes, The control device 300 is By controlling the operation of at least one of the distributed feeder 120 and the supply device 200 based on the estimated quantity, it is possible to supply the articles to the linear feeder 130 that is running low on articles. By configuring it in this way, the amount of goods supplied to the weighing hopper 150 can be optimally optimized by supplying goods to the straight feeder 130 when goods are insufficient.

[0106] Furthermore, the learning model is The estimated amounts for each of the multiple sections that divide the linear feeder 130 in the conveying direction can be output. The control device 300 is Based on the estimated quantities of each of section A and section B (or the multiple sections), the operation of at least one of the distributed feeder 120 and the supply device 200 can be controlled to supply the articles to at least one of section A and section B. By configuring it in this way, the amount of goods supplied to the weighing hopper 150 can be effectively optimized by supplying goods to the parts of section A and section B where goods are insufficient.

[0107] The article according to this embodiment is one form of the object to be measured according to the present invention. Furthermore, the weighing hopper 150 according to this embodiment is one form of the weighing unit according to the present invention. Furthermore, the control device 300 according to this embodiment is one form of implementation of the combination calculation unit according to the present invention. Furthermore, the distributed feeder 120, the linear feeder 130, and the supply device 200 according to this embodiment are one form of the conveying device according to the present invention. Furthermore, the linear feeder 130 according to this embodiment is one form of implementation of the conveying feeder according to the present invention. Furthermore, sections A and B according to this embodiment are one form of implementing multiple parts of the present invention. Furthermore, the image data according to this embodiment is one form of implementing the information regarding imaging results according to the present invention. Furthermore, the weight learning data according to this embodiment is one form of implementing the weight-related information according to the present invention. Furthermore, the learning model generation unit 500 according to this embodiment is one form of the physical simulation unit according to the present invention.

[0108] 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.

[0109] For example, the shapes of the parts of the weighing system 1 described in this embodiment are just examples and are not limited to the shapes described above. The shapes of the parts can be changed to any shape.

[0110] Furthermore, in this embodiment, an example was shown in which a straight feeder 130 is used as a conveying feeder, which is equipped with a straight trough 131 extending along the radial direction so as to move the articles straight outwards radially from the base body 110. However, the conveying feeder is not limited to the example described above, and its shape can be changed as appropriate depending on the direction in which the articles are conveyed.

[0111] Furthermore, although this embodiment shows an example with 14 weighing units (linear feeders 130, supply hoppers 140, and weighing hoppers 150), the number of weighing units is not limited to the example described above, and various numbers can be set. If the number of weighing units is changed, the number of linear feeders 130 in the physical simulation will also be changed.

[0112] Furthermore, although this embodiment describes an example in which both transport control processing and supply control processing can be executed in the model weighing process, the embodiment is not limited to this configuration, and only one of the transport control processing or supply control processing may be executed.

[0113] Furthermore, although this embodiment shows an example in which the linear feeder 130 is divided into two sections, section A and section B, the embodiment is not limited to this configuration, and the linear feeder 130 may be divided into three or more sections.

[0114] Furthermore, although this embodiment shows an example in which granular articles (e.g., food) are used as the object to be weighed, the object to be weighed is not limited to those described above and various other articles can be used. [Explanation of symbols]

[0115] 1. Weighing System 100 Combination Scales 200 Feeding device 300 Control device 400 Imaging device 500 Learning Model Generation Unit 600 Learning Model Execution Unit

Claims

1. Multiple weighing units are arranged along the circumference in a plan view and are capable of weighing the supplied object to be weighed, A combination calculation unit capable of performing combination calculations on the objects to be measured using the measured values ​​of the objects to be measured measured by multiple weighing units, In the transport path from the source of the object to be weighed to the weighing unit, a transport device for transporting the object to be weighed is provided, A control device capable of controlling the operation of the transport device based on the imaging results of the object to be weighed along the transport path, Equipped with, Weighing system.

2. The aforementioned conveying device includes: A distribution feeder that distributes the supplied weighed material horizontally, Multiple conveying feeders transport the objects to be weighed, which have been dispersed by the aforementioned distribution feeder, to a plurality of the aforementioned weighing units. It includes, The control device is Based on the imaging results of the objects to be weighed loaded on the conveyor feeder, the operation of the conveying device is controlled based on the estimated amount of the objects to be weighed on the conveyor feeder. The weighing system according to claim 1.

3. The aforementioned estimate is, The output is generated by a learning model that learns the relationship between the information regarding the imaging results of the object being measured and the information regarding the weight of the object being measured. The weighing system according to claim 2.

4. The system includes a physical simulation unit that obtains information regarding the imaging results of the object to be measured and information regarding the weight of the object to be measured through physical simulation. The weighing system according to claim 3.

5. The aforementioned conveying device includes: The aforementioned conveying feeder is included, The control device is By controlling the operation of the conveying feeder based on the estimated amount, the amount of the object to be weighed supplied from the conveying feeder to the downstream side of the conveying path can be adjusted. The weighing system according to claim 4.

6. The aforementioned learning model, The transport feeder can output the estimated quantities for each of the multiple sections that divide it in the transport direction. The control device is The operation of the transport feeder can be controlled based on the estimated values ​​of each of the multiple parts. The weighing system according to claim 5.

7. The aforementioned conveying device includes: The aforementioned distributed feeder, A supply device that supplies the object to be weighed from the supply source to the distributed feeder, It includes, The control device is By controlling the operation of at least one of the distribution feeder and the supply device based on the estimated quantity, it is possible to supply the weighed material to the conveying feeder when the weighed material is insufficient. The weighing system according to claim 4.

8. The aforementioned learning model, The transport feeder can output the estimated quantities for each of the multiple sections that divide it in the transport direction. The control device is The operation of at least one of the distributed feeder and the supply device can be controlled to supply the weighed object to at least one of the multiple parts based on the estimated quantity of each of the multiple parts. The weighing system according to claim 7.

Images