JP7524600B2 - Box selection device, box selection method, and box selection program - Google Patents

Description

The present invention relates to a box selection device, a box selection method, and a box selection program.

Conventionally, a box selection device such as that described in Patent Document 1 is known. The box selection device selects a combination of cargo and container based on data such as the volume of the target cargo. This box selection device selects all combinations of containers in which the total capacity of the containers is equal to or greater than the total volume of the cargo.

JP 2016-88630 A

The box selection device described above selects all combinations of containers (boxes) that satisfy the conditions. However, when this method is adopted, as the number of types of boxes increases, the number of combinations increases rapidly, and it becomes necessary to calculate each time whether all items can be loaded for all combinations. In this case, there is a problem that adopting this method becomes difficult from the perspective of calculation costs, etc. Furthermore, such calculation costs also increase as the number of types of items increases. In other words, there is a problem that it becomes difficult to apply the box selection device described above to sites where there are a large number of types of items or a large number of box options.

The object of the present invention is to provide a box selection device, a box selection method, and a box selection program that can easily and appropriately select a suitable box when selecting a box to pack an item.

A box selection device according to one aspect of the present invention is a box selection device that selects boxes for packaging items based on previously learned learning information, and the control unit of the box selection device has a storage unit that stores the learning information and box information, an input information receiving unit that creates input information related to the items, a neural network calculation unit that inputs the input information to an input layer and calculates an evaluation value related to the box by neural network calculation based on the previously learned learning information based on packaging performance, and a selection unit that selects a box based on the evaluation value output from the output layer of the neural network calculation unit.

In the box selection device, the neural network calculation unit inputs input information about the item to the input layer, and calculates an evaluation value about the box by neural network calculation based on the learning information. Then, the selection unit can select a box suitable for the item to be packed based on the evaluation value output from the output layer of the neural network calculation unit. Here, the learning information is learned in advance based on the packing record. That is, the neural network calculation unit can omit consideration of complex conditions such as the shape of the item and the environmental information of the packing destination, and can output an evaluation value for selecting a box suitable for the item to be packed by simply learning based on the past packing record. Therefore, the box selection device can easily select a box with reduced calculation cost. In addition, the task of selecting a box suitable for the item to be packed is a task that requires skill for the worker. In contrast, the box selection device can select a box after learning based on the past packing record, so that it can appropriately select a box without being influenced by the skill level of the worker. As described above, the box selection device can easily and appropriately select a suitable box when selecting a box to pack an item.

The input information receiving unit may acquire identification information associated with each of a plurality of items, the input information having a first information table with predetermined rows and columns, the input information receiving unit may input the identification information to the first information table, and the neural network calculation unit and the selection unit may calculate the evaluation value and select the box based on the first information table, respectively. In this case, if there is regularity in the identification information of the items (e.g., product numbers, etc.), it is easier to input the identification information to the neural network calculation unit.

The identification information is managed for each item as a string of a predetermined number of characters, and the input information receiving unit may perform a matching process on the first information table, discarding the excess characters if the number of characters in a row or column exceeds the limit, and filling in the missing parts with a predetermined character if there are insufficient characters in a row or column. In this case, even if the strings of characters for some items are irregular, the input information receiving unit can input the identification information of the items to the neural network calculation unit in a state where it is matched with the first information table.

The input information receiving unit may organize the input order of the identification information to the first information table according to a predetermined rule. In this case, the input information receiving unit can input the identification information in an organized state to the neural network calculation unit.

The input information may have a second information table having rows or columns corresponding to the number of types of items to be packed, and the input information receiving unit may input the number of each item into the second information table, and the neural network calculation unit and the selection unit may calculate the evaluation value and select the box based on the second information table, respectively. In this case, for example, when the item identification information has low regularity, the input information receiving unit may smoothly input the item identification information to the neural network calculation unit.

The output layer may output an evaluation value for each of the multiple boxes in the box information stored in the storage unit. In this case, the selection unit can quickly select one of the boxes stored in the storage unit based on the evaluation value.

The output layer may output the dimensions of each side of the box as an evaluation value. In this case, the selection unit can select a box whose dimensions are close to the evaluation value. Since the selection unit can obtain the appropriate box dimensions as continuous values, this is effective in cases where a box other than those stored in the memory unit is selected.

The neural network calculation unit may calculate the evaluation value, for example, by using a softmax function. In this case, the neural network calculation unit can appropriately calculate the evaluation value.

The neural network calculation unit may calculate the evaluation value using a proportional function. In this case, the neural network calculation unit can appropriately calculate the evaluation value.

A box selection method according to one aspect of the present invention is a box selection method for selecting a box for packaging an item based on previously learned learning information, and includes a storage step for storing the learning information and box information, an input information receiving step for creating input information related to the item, a neural network calculation step for inputting the input information into an input layer and calculating an evaluation value related to the box by neural network calculation based on the previously learned learning information based on packaging performance, and a selection step for selecting a box based on the evaluation value output from the output layer in the neural network calculation step.

A box selection program according to one aspect of the present invention is a box selection program that selects a box for packaging an item based on previously learned learning information, and causes a computer system to execute the following steps: a storage step for storing the learning information and box information; an input information receiving step for creating input information related to the item; a neural network calculation step for inputting the input information into an input layer and calculating an evaluation value related to the box by neural network calculation based on the previously learned learning information based on packaging performance; and a selection step for selecting a box based on the evaluation value output from the output layer in the neural network calculation step.

These box selection methods and box selection programs can achieve the same effects and aims as the box selection device described above.

The present invention provides a box selection device, a box selection method, and a box selection program that can easily and appropriately select a suitable box when selecting a box to pack an item.

1 is a schematic configuration diagram showing a box selection device according to an embodiment of the present invention; 1 is a schematic diagram showing a processing image when the box selection device selects a shipping box using a neural network. FIG. 1 is a schematic diagram showing a processing image when a neural network calculation unit of a box selection device learns using past packing results as training data. FIG. 11 is a schematic diagram showing a specific example of the processing content of the box selection device; FIG. 13 is a diagram showing a process performed when an input information receiving unit inputs identification information into a first information table. FIG. 10 is a flowchart showing the contents of a control process performed by a control unit of the box selection device. 13 is a schematic diagram showing a specific example of the processing content of the box selection device according to the modified example. FIG.

Embodiments of the present invention will be described in detail below with reference to the drawings. In the drawings, identical or equivalent elements are given the same reference numerals, and duplicate descriptions will be omitted.

Figure 1 is a schematic diagram showing a box selection device 100 according to one embodiment of the present invention. The box selection device 100 is a device that selects which box to pack multiple items into when packing them into boxes at a logistics site such as a warehouse. After a box is selected, the position and order in which the items are packed in the box are automatically determined by another method. In this embodiment, parts are exemplified as the items to be packed, and a shipping box is exemplified as the box. However, the types of items and boxes are not particularly limited.

As shown in FIG. 1, the box selection device 100 includes an information input unit 2, an information output unit 3, and a control unit 10. The box selection device 100 is a device that selects a shipping box in which to pack parts based on previously learned learning information. Note that in order to explain each component of the box selection device 100, FIGS. 2 and 3 may be referred to. FIG. 2 is a schematic diagram showing a processing image when the box selection device 100 selects a shipping box using a neural network. FIG. 3 is a schematic diagram showing a processing image when the neural network calculation unit 12 of the box selection device 100 learns using past packaging results as training data.

The information input unit 2 is an interface for inputting various information. For example, the information input unit 2 is composed of a keyboard, a mouse, a touch panel, etc. Alternatively, the information input unit 2 may input information via a storage medium or communication. The information output unit 3 is an interface for outputting various information. For example, the information output unit 3 may output information via a monitor, a speaker, etc. Alternatively, the information output unit 3 may output information via a storage medium or communication.

The control unit 10 is composed of a CPU, RAM, ROM, etc. The control unit 10 has an input information receiving unit 11, a neural network calculation unit 12, a selection unit 13, and a memory unit 14. The control unit 10 can be configured as one or more processors that operate according to a program, one or more dedicated hardware circuits such as ASICs, or a circuit that includes a combination of these. The processor has a CPU, and memories such as RAM and ROM. Various programs for processing information are stored in the memory, and the CPU reads out the various programs and performs calculations.

The input information receiving unit 11 creates input information related to parts. The input information receiving unit 11 acquires information input by the information input unit 2 and creates input information using the acquired information. The input information receiving unit 11 acquires identification information associated with each of a plurality of parts, information related to the number of parts, and the like. The input information is information related to parts that has been adjusted to a form that makes it easy to perform neural network calculations. As shown in FIG. 2, the input information receiving unit 11 acquires a parts list of parts to be packed, i.e., a list showing which parts and how many of each part are to be packed, in the pre-packing stage. Also, as shown in FIG. 3, the input information receiving unit 11 acquires a parts list of parts that have actually been packed and information on which shipping boxes have been used for packing during learning. The specific contents of the identification information associated with the parts and the specific contents of the input information will be described later.

The neural network calculation unit 12 calculates an evaluation value for the shipping box by neural network calculation based on learning information previously learned based on packaging records. The neural network calculation unit 12 acquires learning information by learning using teacher data based on past packaging records (see Figure 3). In addition, the neural network calculation unit 12 uses the learning information to calculate an evaluation value for selecting which shipping box is suitable for the parts list to be packed at the actual work site (see Figure 2).

The neural network calculation unit 12 has a learning unit 16. The learning unit 16 generates a machine learning model as learning information. In the learning stage (see FIG. 3), the learning unit 16 generates the best neural network estimated to have the highest prediction accuracy as a model. The learning unit 16 reads out teacher data (samples of past packaging performance) from the memory unit 14 or the like, and executes machine learning while sequentially inputting each sample to the neural network to be trained. Note that such a neural network can be configured, for example, by a calculation unit or a memory unit that imitates a neuron model. For example, the learning unit 16 may generate a model by executing deep learning using a multilayer neural network. The type of machine learning is not limited to deep learning, and the learning unit 16 may generate a model using other methods. Note that the specific configuration and processing contents of the neural network calculation unit 12 will be described later.

The selection unit 13 selects a shipping box based on the evaluation value output from the output layer of the neural network calculation unit 12. The selection unit 13 acquires the evaluation value and uses the evaluation value to select the most suitable shipping box for packaging the parts. The selection unit 13 selects the most suitable shipping box from a shipping box list (box information) stored in advance in the storage unit 14. The specific processing content of the selection unit 13 will be described later.

The memory unit 14 is a part that stores various information in the box selection device 100. The memory unit 14 stores at least learning information (neural network model) and box information. The box information is information about multiple transport boxes used at the work site, and is stored in the memory unit 14 in advance at a stage prior to the box selection work. The memory unit 14 stores the box information in the form of a transport box list (see Figure 2) that includes information such as the size and shape of each of the multiple transport boxes.

Next, a specific example of the processing of the neural network calculation unit 12 will be described with reference to FIG. 2 and FIG. 3. First, with reference to FIG. 3, a case will be described in which the neural network calculation unit 12 of the box selection device 100 learns using past packing records as teacher data. Here, data is used that indicates that "transport box Z" was used in a parts list of "2 parts a," "1 part b," "2 parts c," and "1 part d" as the packing record. Note that the parts list in this case may be referred to as "parts list PL1." Also, the transport boxes that can be used for packing are selected from a transport box list in which six types of transport boxes, "transport boxes U, V, W, X, Y, and Z" (see FIG. 2), are registered. The transport box list may be referred to as "transport box list BL1."

As shown in FIG. 3, the neural network calculation unit 12 extracts the packed parts list PL1 from the packing performance data, and extracts that "shipping box Z" was selected as the packed shipping box for that parts list. The learning unit 16 (see FIG. 1) of the neural network calculation unit 12 appropriately updates the neural network model M so that when the parts list PL1 is input to the model M, an evaluation value is output that will result in the selection of shipping box Z. The neural network calculation unit 12 learns using multiple types of parts lists and training data based on numerous packing performance records regarding the shipping boxes selected for the parts lists.

Next, referring to FIG. 2, a case where the box selection device 100 selects a transport box will be described. As shown in FIG. 2, the box selection device 100 acquires a transport box list BL1 in advance. Then, when selecting a transport box, the box selection device 100 acquires the above-mentioned parts list PL1 as information indicating the parts to be packed. The neural network calculation unit 12 calculates an evaluation value for the transport box using the neural network model M updated by learning shown in FIG. 3. As a result of learning, the neural network calculation unit 12 outputs an evaluation value that makes transport box Z most likely to be selected from the transport box list BL1. Therefore, the selection unit 13 selects transport box Z based on the evaluation value. The box selection device 100 outputs a message that transport box Z has been selected.

Next, the detailed configuration of the neural network calculation unit 12 and the processing content thereof will be described with reference to FIG. 4. FIG. 4 is a schematic diagram showing a specific example of the processing content of the box selection device 100. First, the processing when the input information receiving unit 11 creates input information will be described. In this embodiment, the input information has a first information table 50 having predetermined rows and columns. As a result, the neural network calculation unit 12 and the selection unit 13 calculate the evaluation value and select the shipping box, respectively, based on the first information table 50.

The first information table 50 shown in FIG. 4 has storage sections 51 with "N rows x M columns". Each storage section 51 stores one character. The number of characters per row in the first information table 50 is fixed at M. The number of rows in the first information table 50 is also fixed at N. In contrast, the identification information for each part is composed of a character string indicating a part number, etc. The identification information is managed as a character string of a predetermined number of characters for each part. Here, it is preferable that the number of characters in the identification information for each part is set to M, matching the number of characters per row in the first information table 50. In the example shown in FIG. 4, the identification information for "part a" is composed of M numbers "1", i.e., "111...1".

The input information receiving unit 11 inputs the identification information into the first information table 50. In this way, input information is created by storing characters related to the identification information of each part in each storage unit of the first information table 50. Note that the identification information corresponding to one part is stored in one row of the storage unit 51 of the first information table 50. Therefore, if there are N parts in the parts list, the identification information of all parts can be stored in all rows of the first information table 50 without any remainder. Note that the characters indicating the identification information are not limited to numbers, and may be letters of the alphabet or special characters such as symbols as long as they can identify the part.

Here, when the number of characters in the rows and columns of the first information table 50 is exceeded, the input information receiving unit 11 discards the excess characters as a matching process. When the number of characters in the rows and columns of the first information table 50 is insufficient, the input information receiving unit 11 fills the missing parts with a predetermined character as a matching process. Specifically, as shown in FIG. 5(a), the number of columns of the first information table 50, that is, the number of storage sections 51 for one row, is fixed at 6 (M=6). Here, the identification information (product number) of part f is composed of four characters, which is fewer than the number of columns. In this case, a missing part 53 of two storage sections 51 is formed in the row corresponding to part f. In this way, when the number of characters in a column is insufficient for the first information table 50, the input information receiving unit 11 fills the missing part 53 with a predetermined character (here, the number "0") as a matching process. On the other hand, the identification information (product version) of part g is composed of eight characters, which is more than the number of columns. In this case, two excess points 54 occur in the row corresponding to part g. In this way, when the number of characters in a column exceeds a predetermined number (here, six) for the first information table 50, the input information receiving unit 11 discards the excess characters (the characters in the excess points 54) as a matching process.

As shown in FIG. 5(b), the number of rows in the first information table 50, that is, the number of storage sections 51 in one column, is fixed at 4 (N=4). Here, the number of parts in the parts list is 3, which is less than the number of rows. In this case, a missing portion 56 of the storage section 51 is formed in the fourth row. In this way, when there is a lack of characters in a row for the first information table 50, the input information receiving unit 11 fills the missing portion 56 with a predetermined character (here, the number "0") as a matching process. On the other hand, in FIG. 5(c), the number of parts in the parts list is 5, which is more than the number of rows. In this case, an excess portion 57 occurs in the fifth row. In this way, when the number of characters in a row (here, 4) is exceeded for the first information table 50, the input information receiving unit 11 discards the excess characters (characters in the excess portion 57) as a matching process.

The input information receiving unit 11 may organize the input order of the identification information into the first information table 50 according to a predetermined rule. For example, the input order of the part numbers of the parts to be packed may be organized in ascending or descending order of the part numbers, or, if weight information is available, in order of weight. For example, in FIG. 5(c), the part numbers of parts e to i are each composed of a string of numbers from 4 to 8. In this case, the input information receiving unit 11 inputs parts e to i into the first information table 50 in order from the first row in descending order of part number number.

When it is desired to indicate in the input information that there are multiple identical parts, the input information receiving unit 11 may input the same part number consecutively for the number of lines equal to the number of identical parts. For example, as shown in FIG. 5(d), if there are three parts g, the part number of part g is consecutively input for lines 3 to 5 to form consecutive part number portion 58.

Returning to FIG. 4, the input information receiving unit 11 creates input information by storing the identification information (product number) of each part in the first information table 50, and then inputs the character information stored in each row to the neural network calculation unit 12.

The neural network calculation unit 12 has an input layer 21, an intermediate layer 22, and an output layer 23 as a model of a neural network. The input layer 21 is a layer for inputting parameters for performing neural network calculations. The intermediate layer 22 is a layer for performing a predetermined calculation on the input parameters. The output layer 23 is a layer for outputting an evaluation value. Each of the layers 21 to 23 has nodes ND according to the number of input parameters. For example, the input layer 21 outputs an input vector x=(x ₁ , x ₂ , . . . x _m ) having m parameters as components to the intermediate layer 22 as is. The intermediate layer 22 converts the total input into an output using the weights and activation functions of each layer, and passes the output to the output layer 23. The output layer 23 also converts the total input into an output using the weights and activation functions of each layer. This output is an output vector y=(y ₁ , y ₂ , . . . , y _n ) of the neural network having n parameters as components. This output vector y indicates an evaluation value. During learning (see FIG. 3), if the correct output vector for one input vector x is d=( _d1 , _d2 , ..., _dn ), the weights w of each layer are updated so that the output vector y approaches the correct answer d. This is the information processing called learning.

In this embodiment, the number of nodes ND in the input layer 21 is set to the same as the number of rows (N) in the first information table 50, i.e., the number of parts included in the input information. The intermediate layer 22 is composed of typical layers in neural networks, such as fully connected layers, convolutional layers, and pooling layers, and is appropriately configured according to the work site.

The number of nodes ND in the output layer 23 is set to the same number of selectable transport boxes, that is, the number of transport boxes in the transport box list BL1 (see FIG. 2). Each transport box is assigned to each node ND in the output layer 23, and the output value of the corresponding node ND is set as the evaluation value of each transport box. Here, the neural network calculation unit 12 calculates the evaluation value using a softmax function. Specifically, the output layer 23 calculates the evaluation value using a softmax function. The softmax function is shown in the following formula (1). Note that the softmax function calculates the output of each node ND so that the sum of the outputs of all nodes ND is 1. The output layer 23 sets this output value as the evaluation value of the transfer box assigned to each node ND. As a result, the output layer 23 can output an evaluation value for each of the multiple transport boxes in the box information stored in the storage unit 14. Note that the larger the evaluation value, the more suitable the transport box is for packaging.

The selection unit 13 sets the output value of the neural network calculation unit 12 as the evaluation value of each transport box. For example, the selection unit 13 acquires the output value for transport box U as evaluation value y1, the output value for transport box V as evaluation value y2, and the output value for transport box Z as evaluation value yn. Then, the selection unit 13 selects the transport box with the maximum evaluation value.

Next, the box selection method according to this embodiment will be described with reference to FIG. 6. FIG. 6 is a flowchart showing the contents of the control process by the control unit 10 of the box selection device 100. The box selection method is a method for selecting a shipping box in which to pack parts based on previously learned learning information. Note that the detailed processing contents in each step are the same as those described in the description of each component of the box selection device 100, and therefore will not be described here.

As shown in FIG. 6, a storage step S10 is executed in which the learning information and the box information of the shipping box are stored in the memory unit 14. Note that the storage step S10 may be executed in advance before the process of FIG. 6 is executed. Here, as described in FIG. 3, a neural network model is generated or updated as the learning information.

Next, the input information receiving unit 11 executes an input information receiving step S20 in which it obtains information related to the parts and creates input information based on the obtained information. Next, the neural network calculation unit 12 inputs the input information created in the input information receiving step S20 to the input layer 21, and executes a neural network calculation step S30 in which it calculates an evaluation value related to the shipping box by neural network calculation based on the learning information stored in the memory unit 14. Next, the selection unit 13 executes a selection step S40 in which it selects a shipping box based on the evaluation value output from the output layer 23 in the neural network calculation step S30. With the above, the control process shown in FIG. 6 is completed. When selecting a shipping box for another parts list, the process may be repeated again from the input information receiving step S20.

Next, we will explain the operation and effects of the box selection device 100 according to this embodiment.

The box selection device according to this embodiment is a box selection device 100 that selects a box for packaging an item based on previously learned learning information. The control unit 10 of the box selection device 100 has a memory unit 14 that stores learning information and box information, an input information receiving unit 11 that creates input information related to the item, a neural network calculation unit 12 that inputs the input information to an input layer 21 and calculates an evaluation value related to the box by neural network calculation based on the learning information previously learned based on packaging performance, and a selection unit 13 that selects a box based on the evaluation value output from the output layer 23 of the neural network calculation unit 12.

In the box selection device 100, the neural network calculation unit 12 inputs input information about the item to the input layer 21, and calculates an evaluation value for the box by neural network calculation based on the learning information. Then, the selection unit 13 can select a box suitable for the item to be packed based on the evaluation value output from the output layer 23 of the neural network calculation unit 12. Here, the learning information is learned in advance based on the packing record. That is, the neural network calculation unit 12 can output an evaluation value for selecting a box suitable for the item to be packed by omitting to consider complex conditions such as the shape of the item and the environmental information of the packing destination, and only by learning based on the past packing record. Therefore, the box selection device 100 can easily select a box with reduced calculation costs. In addition, the task of selecting a box suitable for the item to be packed is a task that requires skill for the worker. In contrast, the box selection device 100 can select a box after learning based on the past packing record, so that the box can be appropriately selected without being influenced by the worker's skill level. As a result, the box selection device 100 can easily and appropriately select a suitable box when selecting a box to pack an item.

The input information receiving unit 11 acquires identification information linked to each of a plurality of items, and the input information has a first information table 50 having predetermined rows and columns. The input information receiving unit 11 inputs the identification information to the first information table 50, and the neural network calculation unit 12 and the selection unit 13 may calculate the evaluation value and select the box, respectively, based on the first information table 50. In this case, if there is regularity in the identification information of the items (e.g., product number, etc.), it becomes easier to input the identification information to the neural network calculation unit 12.

The identification information is managed for each item as a string of a predetermined number of characters, and the input information receiving unit 11 may perform a matching process on the first information table 50, discarding the excess characters if the number of characters in a row or column exceeds the limit, and filling in the missing parts with a predetermined character if the number of characters in a row or column is insufficient. In this case, even if the strings of characters for some items are irregular, the input information receiving unit 11 can input the identification information of the items to the neural network calculation unit 12 in a state that is matched with the first information table 50.

The input information receiving unit 11 may organize the input order of the identification information to the first information table 50 according to a predetermined rule. In this case, the input information receiving unit 11 can input the identification information in an organized state to the neural network calculation unit 12, suppressing the effect of the input order of the input information on the selection result of the transport box, and improving the selection accuracy.

The output layer 23 may output an evaluation value for each of the multiple boxes in the box information stored in the memory unit 14. In this case, the selection unit 13 can quickly select one of the boxes stored in the memory unit 14 based on the evaluation value.

The neural network calculation unit 12 may calculate the evaluation value using a softmax function. In this case, the neural network calculation unit 12 can appropriately calculate the evaluation value.

The box selection method according to this embodiment is a box selection method for selecting a box for packaging an item based on previously learned learning information, and includes a storage step S10 for storing the learning information and box information, an input information receiving step S20 for creating input information related to the item, a neural network calculation step S30 for inputting the input information into the input layer 21 and calculating an evaluation value related to the box by neural network calculation based on the previously learned learning information based on packaging performance, and a selection step S40 for selecting a box based on the evaluation value output from the output layer 23 in the neural network calculation step S30.

The box selection program according to this embodiment is a box selection program that selects a box for packaging an item based on previously learned learning information, and causes a computer system to execute a storage step S10 for storing the learning information and box information, an input information receiving step S20 for creating input information related to the item, a neural network calculation step S30 for inputting the input information into an input layer 21 and calculating an evaluation value related to the box by neural network calculation based on previously learned learning information based on packaging performance, and a selection step S40 for selecting a box based on the evaluation value output from the output layer 23 in the neural network calculation step S30.

These box selection methods and box selection programs can achieve the same effects and benefits as the box selection device 100 described above.

The present invention is not limited to the above-described embodiments.

For example, the box selection device 100 may perform a process as shown in FIG. 7. In the modified example shown in FIG. 7, the input information has a second information table 60 having rows (or columns) corresponding to the number of types of parts to be packed. That is, when there are N types of parts, the second information table 60 has N rows of storage sections 61. Note that there is only one storage section 61 in each row (i.e., one column). The input information receiving unit 11 inputs the number of parts related to each type to the second information table 60. For example, in the example shown in FIG. 7, "0 parts a, 2 parts b, 0 parts c, ... 1 part z". Therefore, "2" is stored in the storage section 61 in the second row corresponding to part b in the second information table 60, and "1" is stored in the storage section 61 in the Nth row corresponding to part z. The neural network calculation unit 12 and the selection unit 13 calculate the evaluation value and select the shipping box based on the second information table 60, respectively.

When there are "unknown parts" whose details are not yet known, such as new parts, the input information receiving unit 11 may perform the following preprocessing. For example, the input information receiving unit 11 may add a row indicating the "unknown parts" to the second information table 60, and input the number of all unknown parts in that row. Then, the input information receiving unit 11 may discard the row corresponding to the "unknown parts".

In the modified example shown in FIG. 7, the number of nodes ND in the input layer 21 of the neural network calculation unit 12 is the same as the number of rows in the second information table 60, i.e., the number of all part types. The neural network calculation unit 12 calculates the evaluation value using a proportional function. Specifically, the output layer 23 calculates the evaluation value using a proportional function. The proportional function is shown in the following formula (2). The output layer 23 outputs the dimensions of each side of the shipping box as the evaluation value. That is, the long side, short side, and height of the shipping box are assigned to each node ND of the output layer 23, and the dimensions of each side are output.

yk = zk ... (2)

yk: Output of the kth node in the output layer zk: Value of the kth node in the output layer

The selection unit 13 obtains the dimensions of each side as the evaluation value output from the output layer 23. Then, the selection unit 13 compares the dimensions of each side of the selectable transport boxes (transport boxes in the transport box list BL1) with the dimensions related to the evaluation value to calculate an error. Then, the selection unit 13 selects a transport box that has a minimum error with the dimensions related to the evaluation value. Here, the selection unit 13 may use the sum of squares of the errors of each side (see the following formula (3)) as an evaluation function during selection.

As in the modified example shown in FIG. 7, the input information has a second information table 60 having rows or columns corresponding to the number of types of items to be packed, and the input information receiving unit 11 inputs the number of each item into the second information table 60, and the neural network calculation unit 12 and the selection unit 13 may calculate the evaluation value and select the box based on the second information table 60, respectively. In this case, for example, when the item identification information has low regularity, the input information receiving unit 11 can smoothly input the item identification information to the neural network calculation unit 12.

The neural network calculation unit 12 may calculate the evaluation value using a proportional function. In this case, the neural network calculation unit 12 can appropriately calculate the evaluation value.

The output layer 23 may output the dimensions of each side of the box as the evaluation value. In this case, the selection unit 13 can select a box whose dimensions are close to the evaluation value. The selection unit 13 can obtain the appropriate box dimensions as continuous values, which is effective in cases where a box other than those stored in the memory unit 14 is selected.

The processing of the output layer 23 described in the embodiment shown in FIG. 4 and the processing of the output layer 23 described in the modified example shown in FIG. 7 may be interchanged.

10...control unit, 11...input information receiving unit, 12...neural network calculation unit, 13...selection unit, 14...storage unit, 21...input layer, 23...output layer, 100...box selection device.

Claims (7)

A box selection device that selects a box for packaging an item based on previously learned learning information,
The control unit of the box selection device
A storage unit that stores the learning information and box information;
an input information receiving unit for creating input information regarding the item;
a neural network calculation unit that inputs the input information into an input layer and calculates an evaluation value for the box by neural network calculation based on the learning information that has been learned in advance based on packaging records;
a selection unit that selects the box based on the evaluation value output from an output layer of the neural network calculation unit,
The output layer outputs the value of each node and the dimension of each side of the box as the evaluation value. The input information receiving unit acquires identification information associated with each of the plurality of articles,
the input information comprises a first table of information having predetermined rows and columns;
the input information receiving unit inputs the identification information into the first information table;
2. The box selection device according to claim 1 , wherein the neural network calculation unit and the selection unit calculate the evaluation value and select the box, respectively, based on the first information table. A box selection device that selects a box for packaging an item based on previously learned learning information,
The control unit of the box selection device
A storage unit that stores the learning information and box information;
an input information receiving unit for creating input information regarding the item;
a neural network calculation unit that inputs the input information into an input layer and calculates an evaluation value for the box by neural network calculation based on the learning information that has been learned in advance based on packaging records;
a selection unit that selects the box based on the evaluation value output from an output layer of the neural network calculation unit,
The input information receiving unit acquires identification information associated with each of the plurality of articles,
the input information comprises a first table of information having predetermined rows and columns;
the input information receiving unit inputs the identification information into the first information table;
the neural network calculation unit and the selection unit calculate the evaluation value and select the box, respectively, based on the first information table;
The identification information is managed for each item as a character string of a predetermined number of characters,
The input information receiving unit, with respect to the first information table,
If the number of characters in a row or column is exceeded, the excess characters are discarded;
If there are insufficient characters in a row or column, the box selection device performs a matching process to fill in the insufficient portions with a predetermined character. 4. The box selection device according to claim 2, wherein the input information receiving unit arranges an input order of the identification information to the first information table according to a predetermined rule. A box selection device that selects a box for packaging an item based on previously learned learning information,
The control unit of the box selection device
A storage unit that stores the learning information and box information;
an input information receiving unit for creating input information regarding the item;
a neural network calculation unit that inputs the input information into an input layer and calculates an evaluation value for the box by neural network calculation based on the learning information that has been learned in advance based on packaging records;
a selection unit that selects the box based on the evaluation value output from an output layer of the neural network calculation unit,
The input information includes a second information table having rows or columns corresponding to the number of types of items to be packed,
the input information receiving unit inputs the number of each of the items into the second information table;
The neural network calculation unit and the selection unit calculate the evaluation value and select the box, respectively, based on the second information table. A box selection method for selecting a box for packaging an item based on previously learned learning information, comprising:
a storage step of storing the learning information and the box information;
an input information receiving step for creating input information regarding the item;
a neural network calculation step of inputting the input information into an input layer and calculating an evaluation value for the box by neural network calculation based on the learning information that has been learned in advance based on packaging records;
a selection step of selecting the box based on the evaluation value output from an output layer in the neural network operation step,
In the neural network calculation step, the evaluation value is calculated using a softmax function;
the output layer outputs the evaluation value for each of the plurality of boxes in the box information stored in the storing step;
A box selection method , in which the output layer outputs the value of each node and the dimensions of each side of the box as the evaluation value. A box selection program for selecting a box for packaging an item based on previously learned learning information,
a storage step of storing the learning information and the box information;
an input information receiving step for creating input information regarding the item;
a neural network calculation step of inputting the input information into an input layer and calculating an evaluation value for the box by neural network calculation based on the learning information that has been learned in advance based on packaging records;
a selection step of selecting the box based on the evaluation value output from the output layer in the neural network operation step;
In the neural network calculation step, the evaluation value is calculated using a softmax function;
the output layer outputs the evaluation value for each of the plurality of boxes in the box information stored in the storing step;
The output layer outputs the value of each node and the dimensions of each side of the box as the evaluation value.

Images