Estimation device, estimation method, estimation system, determination method, and computer program

Abstract

This estimation device comprises a control unit that, by using an estimation model constructed in advance by using known data including supplementary information that is information on iron scrap including iron to be recycled and information on a concentration of an impurity element included in the iron scrap, estimates a concentration of an impurity element included in iron scrap to be evaluated.

Description

[Technical Field]

[0001] The present invention relates to an estimation device, an estimation method, an estimation system, a determination method, and a computer program. This application claims priority to Japanese Patent Application No. 2023-181202, filed on October 20, 2023, the contents of which are incorporated herein by reference. [Background technology]

[0002] In recent years, there has been an increasing demand in the steel industry for the recycling of steel, with the aim of reducing carbon dioxide emissions and other benefits. There are several types of scrap (steel scrap) that contain iron that can be recycled. Scrap that is generated mainly in the processing of steel products is classified as high-grade scrap, while H2, which is generated mainly from building demolition, is classified as low-grade scrap. To manufacture high-grade steel, it is desirable to use high-grade high-grade scrap, but using only high-grade scrap could lead to future supply shortages. Therefore, it is necessary to utilize low-grade scrap as well.

[0003] Low-grade scrap contains a high percentage of impurity elements other than iron, and the content varies widely. This is largely due to the inclusion of prohibited items such as motors and power distribution boards in the scrap. Among impurity elements, elements that are difficult to remove during the steelmaking process (tramp elements: TE, representative elements: Cu, Sn, Ni, Cr, and Mo) can cause cracks during hot rolling and lead to poor material properties (mechanical properties). To address these issues, technologies have been proposed for measuring the concentration of specific impurity elements after melting (see, for example, Patent Documents 1 and 2). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2007-322190 [Patent Document 2] Japanese Patent Application Laid-Open No. 2011-168823 Summary of the Invention [Problem to be solved by the invention]

[0005] However, with conventional post-dissolution concentration measurement techniques, it is difficult to remove TE even if it is incorporated at levels above the standard. Therefore, the present invention has been made in consideration of the above-mentioned circumstances, and provides a technology that makes it possible to more accurately obtain the TE components in iron to be recycled. [Means for solving the problem]

[0006] (1) One aspect of the present invention is an estimation device that includes a control unit that estimates the concentration of impurity elements contained in iron scrap to be evaluated using an estimation model that is pre-constructed using known data including supplementary information that is information about iron scrap containing iron to be recycled and information about the concentration of impurity elements contained in the iron scrap.

[0007] (2) One aspect of the present invention is the estimation device of (1) above, wherein the supplementary information includes attribute information of the iron scrap itself or information about an inspection performed in advance on the iron scrap.

[0008] (3) One aspect of the present invention is the estimation device according to (1) or (2) above, wherein the supplementary information includes information about the appearance of the iron scrap.

[0009] (4) One aspect of the present invention is the estimation device according to (1) above, wherein the supplementary information includes information about prohibited substances contained in the iron scrap.

[0010] (5) One aspect of the present invention is the estimation device of (4) above, wherein the information on the prohibited substances contained in the iron scrap indicates the number or types of prohibited substances estimated by using a prohibited substance estimation model used to estimate prohibited substances that has been constructed in advance.

[0011] (6) One aspect of the present invention is an estimation method having an estimation step of estimating the concentration of impurity elements contained in the iron scrap to be evaluated using an estimation model constructed in advance using known data including supplementary information, which is information about the iron scrap containing iron to be recycled, and information about the concentration of impurity elements contained in the iron scrap.

[0012] (7) One aspect of the present invention is a computer program for causing a computer to function as an estimation device that includes a control unit that estimates the concentration of impurity elements contained in iron scrap to be evaluated using an estimation model that is pre-constructed using known data including supplementary information that is information about iron scrap containing iron to be recycled and information about the concentration of impurity elements contained in the iron scrap.

[0013] (8) One aspect of the present invention is an estimation system that includes an estimation unit that estimates the concentration of impurity elements contained in the iron scrap to be evaluated using an estimation model that is pre-constructed using known data including supplementary information, which is information about iron scrap containing iron to be recycled, and information about the concentration of impurity elements contained in the iron scrap, and a removal target determination unit that determines taboo materials that should be removed from the iron scrap based on the estimation results by the estimation unit.

[0014] (9) One aspect of the present invention is the estimation system of (8) above, wherein the supplementary information includes information about the appearance of the iron scrap.

[0015] (10) One aspect of the present invention is a determination method including an estimation step of estimating the concentration of impurity elements contained in the iron scrap to be determined using an estimation model previously constructed using known data including supplementary information, which is information about iron scrap containing iron to be recycled, and information about the concentration of impurity elements contained in the iron scrap, and a removal target determination step of determining taboo materials to be removed from the iron scrap based on the estimation results in the estimation step.

[0016] (11) One aspect of the present invention is a computer program for causing a computer to function as an estimation system including: an estimation unit that estimates the concentration of impurity elements contained in the iron scrap to be evaluated using an estimation model previously constructed using known data including supplementary information, which is information about iron scrap containing iron to be recycled, and information about the concentration of impurity elements contained in the iron scrap; and a removal target determination unit that determines taboo materials to be removed from the iron scrap based on the estimation results by the estimation unit.

[0017] (12) One aspect of the present invention is an estimation system that includes an estimation unit that estimates the concentration of impurity elements contained in the iron scrap to be evaluated using an estimation model that is pre-constructed using known data including supplementary information, which is information about iron scrap containing iron to be recycled, and information about the concentration of impurity elements contained in the iron scrap, and a blending information determination unit that determines the blending of the iron scrap to satisfy specified conditions based on the estimation result by the estimation unit and the price of the iron scrap.

[0018] (13) One aspect of the present invention is a determination method including: an estimation step of estimating the concentration of impurity elements contained in the iron scrap to be determined using an estimation model constructed in advance using known data including supplementary information, which is information about iron scrap containing iron to be recycled, and information about the concentration of impurity elements contained in the iron scrap; and a composition information determination step of determining the composition of the iron scrap to satisfy specified conditions based on the estimation result in the estimation step and the price of the iron scrap.

[0019] (14) One aspect of the present invention is a computer program for causing a computer to function as an estimation system including: an estimation unit that estimates the concentration of impurity elements contained in the iron scrap to be evaluated using an estimation model previously constructed using known data including supplementary information, which is information about iron scrap containing iron to be recycled, and information about the concentration of impurity elements contained in the iron scrap; and a composition information determination unit that determines the composition of the iron scrap to satisfy specified conditions based on the estimation result by the estimation unit and the price of the iron scrap.

[0020] (15) One aspect of the present invention is an estimation system that includes an estimation unit that estimates the concentration of impurity elements contained in iron scrap to be evaluated using an estimation model that is pre-constructed using known data including supplementary information that is information about iron scrap containing iron to be recycled and information about the concentration of impurity elements contained in the iron scrap, and a price determination unit that determines the price of the iron scrap based on the estimation result by the estimation unit.

[0021] (16) One aspect of the present invention is a determination method comprising an estimation step of estimating the concentration of impurity elements contained in the iron scrap to be determined using an estimation model constructed in advance using known data including supplementary information, which is information about iron scrap containing iron to be recycled, and information about the concentration of impurity elements contained in the iron scrap, and a price determination step of determining the price of the iron scrap based on the estimation result in the estimation step.

[0022] (17) One aspect of the present invention is a computer program for causing a computer to function as an estimation system including: an estimation unit that estimates the concentration of impurity elements contained in the iron scrap to be evaluated using an estimation model previously constructed using known data including supplementary information, which is information about iron scrap containing iron to be recycled, and information about the concentration of impurity elements contained in the iron scrap; and a price determination unit that determines the price of the iron scrap based on the estimation result by the estimation unit. [Effects of the Invention]

[0023] According to the present invention, it is possible to obtain the concentration of impurity elements in iron to be recycled with higher accuracy. [Brief explanation of the drawings]

[0024] [Figure 1] 1 is a schematic block diagram showing a system configuration of a first embodiment of an estimation system 100 of the present invention. [Figure 2] FIG. 1 is a diagram illustrating an application example of an estimation system 100 according to an embodiment of the present invention. [Figure 3] 2 is a schematic block diagram showing a specific example of the functional configuration of the terminal device 10. FIG. [Figure 4] 2 is a schematic block diagram showing a specific example of the functional configuration of an estimation device 30. FIG. [Figure 5] FIG. 10 is a diagram illustrating a first specific example of training data. [Figure 6] FIG. 6 is a diagram showing an example of an estimation model obtained by learning using the training data shown in FIG. 5. [Figure 7] FIG. 10 is a diagram illustrating a second specific example of training data. [Figure 8] FIG. 8 is a diagram showing an example of an estimation model obtained by learning using the training data shown in FIG. 7. [Figure 9] FIG. 10 is a diagram illustrating a third specific example of training data. [Figure 10] FIG. 10 is a diagram showing an example of an estimation model obtained by learning using the training data shown in FIG. 9. [Figure 11] FIG. 10 is a diagram illustrating a fourth specific example of training data. [Figure 12] FIG. 12 is a diagram showing an example of an estimation model obtained by learning using the training data shown in FIG. 11. [Figure 13] 10 is a flowchart showing a specific example of processing by the estimation device 30. [Figure 14] FIG. 2 is a schematic block diagram showing a specific example of the functional configuration of a model construction device 40. [Figure 15] 10 is a flowchart showing a specific example of processing by the model construction device 40. [Figure 16] FIG. 10 is a diagram illustrating a modified example of the estimation device 30. [Figure 17] FIG. 10 is a diagram illustrating a modified example of the estimation device 30. [Figure 18] FIG. 2 is a schematic block diagram showing a system configuration of a second embodiment of an estimation system 100 according to the present invention. [Figure 19] 2 is a schematic block diagram showing a specific example of the functional configuration of a contraindicated object estimation device 50. FIG. [Figure 20] 10 is a flowchart showing a specific example of the process of the contraindicated object estimation device 50. [Figure 21] FIG. 10 is a diagram illustrating a modified example of the estimation device 30. [Figure 22] FIG. 10 is a schematic block diagram showing a system configuration of a third embodiment of an estimation system 100 according to the present invention. [Figure 23] 1 is a schematic block diagram showing a specific example of the functional configuration of a removal target determination device 60. FIG. [Figure 24] FIG. 10 is a diagram showing a specific example of taboo object information. [Figure 25] 10 is a flowchart showing a specific example of the process of the removal target determination unit 632. [Figure 26] FIG. 1 is a diagram illustrating an application example of an estimation system 100 according to an embodiment of the present invention. [Figure 27] FIG. 10 is a schematic block diagram showing the system configuration of a fourth embodiment of an estimation system 100 according to the present invention. [Figure 28] 1 is a schematic block diagram showing a specific example of the functional configuration of a blending information determination device 70. FIG. [Figure 29] FIG. 10 is a diagram showing a specific example of inventory information. [Figure 30] FIG. 10 is a schematic block diagram showing a system configuration of a fifth embodiment of an estimation system 100 according to the present invention. [Figure 31] 2 is a schematic block diagram showing a specific example of the functional configuration of a price determination device 80. FIG. [Figure 32] FIG. 10 is a diagram showing a specific example of price information. [Figure 33]FIG. 2 is a diagram illustrating an outline of an example of the hardware configuration of an information processing device 90 applied to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0025] [First embodiment] A specific configuration example of the present invention will be described with reference to the drawings. FIG. 1 is a schematic block diagram showing the system configuration of a first embodiment of an estimation system 100 of the present invention. The estimation system 100 is used to estimate the concentration of impurity elements contained in iron scrap to be estimated (hereinafter referred to as "target scrap"). In the following description, copper will be used as a specific example of an impurity contained in the target scrap. However, the types of impurities that can be estimated using the estimation system 100 do not necessarily have to be limited to copper. In the following description, impurities refer to elements other than iron (e.g., copper), and prohibited materials refer to parts in the scrap other than iron (e.g., motors).

[0026] The estimation system 100 includes a terminal device 10, an imaging device 20, an estimation device 30, and a model construction device 40. The terminal device 10, the imaging device 20, and the estimation device 30 are communicatively connected via a network 200. The estimation device 30 and the model construction device 40 may be communicatively connected via the network 200. The network 200 may be a network using wireless communication or a network using wired communication. The network 200 may be configured using, for example, the Internet or a local area network (LAN). The network 200 may be configured by combining multiple networks. The network used when the terminal device 10 and the estimation device 30 communicate may be different from the network used when the imaging device 20 and the estimation device 30 communicate.

[0027] First, an overview of the scrap distribution channel to which the estimation system 100 of this embodiment is applied will be described. First, iron scrap is generated when structures made of iron are dismantled or discarded. The dismantling of the structures may be carried out, for example, by a dismantling contractor. At this time, information about the structure before dismantling or discarding, information about the dismantling contractor, information indicating the status at the time of dismantling, information indicating the status at the time of discarding, and the like may be treated as part of the incidental information.

[0028] The generated scrap iron is collected by collection companies. The collected scrap iron is processed by scrap processing companies. These collection companies and scrap processing companies are collectively called scrap dealers. Scrap dealers may take images of the scrap iron when they collect (receive). The images taken at this time may be treated as part of the additional information. Scrap dealers sort the received scrap iron and transport the sorted scrap iron to scrap users. If the received scrap iron is large, the scrap dealer may process the scrap iron. Examples of such processing include shearing (guillotine, gas), shredding (crushing), and pressing. Processing allows scrap iron to be made into smaller shapes.

[0029] In sorting iron scrap, the estimation system 100 may use estimated values ​​of the tramp element components contained in each iron scrap. By sorting using such estimated values, scrap dealers can predict the concentrations of tramp element components contained in the iron scrap in advance and set and achieve sorting goals that meet the requirements of the purchaser (scrap user). Information about the scrap dealer that handled the iron scrap, information about the processing of the iron scrap, information about the scrap dealer's personnel, and information showing the results of sorting the iron scrap may be treated as part of the incidental information.

[0030] The scrap steel user who receives the transported scrap accepts the transported scrap from the scrap dealer and purchases some or all of the scrap. At this time, images of the scrap steel taken at the time of transport and information about the transporter of the scrap steel may be treated as part of the additional information. The purchase price may be determined based on a predetermined standard. For example, it may be determined based on the shape and size of the scrap steel, or it may be determined based on the tramp element content contained in the scrap steel. The tramp element content may be estimated by the estimation system 100 using the additional information described above. The purchased scrap steel is managed by lot, crushed into small pieces, and the blending ratio is determined. The scrap steel is then melted in an electric furnace. At this time, the purchase price and images taken during crushing may be used as part of the additional information. By using the estimation system 100 to estimate the tramp element content contained in the purchased scrap steel for each lot, scrap users can obtain the desired steel composition.

[0031] Actual measured values ​​of the tramp element components may be obtained by analyzing the components of the dissolved material. These actual measured values ​​may be fed back to the estimation system 100, and the estimation model may be updated using the estimation results and the actual measured values.

[0032] FIG. 2 is a diagram showing an application example of the estimation system 100 of this embodiment (components other than the terminal device 10 and the imaging device 20 are not shown). In the application example shown in FIG. 2, the estimation system 100 is applied to an iron scrap yard, which is a place where iron scrap to be recycled (target scrap) is accumulated. The iron scrap yard is equipped with an imaging device 20 and a conveying device 300. The imaging device 20 is installed at a position where it can capture an image of iron scrap loaded on the bed of a truck transporting the iron scrap on the premises of the above-mentioned dismantler, scrap dealer, iron scrap user, etc. For example, the imaging device 20 may be installed to capture an image of the bed of a truck parked in a predetermined parking position. The conveying device 300 is installed to transport the iron scrap from the bed of the truck to an iron scrap loading area. The conveying device 300 includes, for example, a lift magnet 301, a crane 302, and a crane rail 303.

[0033] Steel scrap generated by demolishing buildings in factories and cities is transported to a steel scrap yard by truck. The truck, with the steel scrap loaded on its bed, is parked in a designated parking spot. When the truck is parked, the imaging device 20 captures an image of the area including the bed and generates image data. The imaging device 20 transmits the generated image data to the estimation device 30. After the imaging device 20 has finished capturing images, the transport device 300 transports the steel scrap from the bed of the truck to a steel scrap loading site.

[0034] At the iron scrap yard, a worker performing work related to loading and unloading of iron scrap operates the terminal device 10 to input inspection performance information of the supplementary information. The inspection performance information includes information about the supplier (e.g., a profile) and information indicating where the iron scrap was generated. The supplier may input the inspection performance information by operating the terminal device 10. The terminal device 10 accepts the input of the inspection performance information in response to the operator's operation and transmits the input inspection performance information to an estimation device 30 (not shown). The estimation device 30 performs estimation processing using supplementary information (including inspection performance information and / or image data) acquired by the above-mentioned dismantler, scrapper, iron scrap user, etc. The estimation device 30 transmits estimation results obtained by the estimation processing to the terminal device 10. Upon receiving the estimation results from the estimation device 30, the terminal device 10 displays the received estimation results. Next, each device used in the estimation system 100 will be described.

[0035] 3 is a schematic block diagram showing a specific example of the functional configuration of the terminal device 10. The terminal device 10 is configured using information equipment such as a smartphone, tablet, personal computer, dedicated device, etc. The terminal device 10 includes a communication unit 11, an operation unit 12, an output unit 13, a storage unit 14, and a control unit 15.

[0036] The communication unit 11 is a communication device. The communication unit 11 may be configured as, for example, a network interface. The communication unit 11 communicates data with other devices via the network 200 in accordance with the control of the control unit 15. The communication unit 11 may be a device that performs wireless communication or a device that performs wired communication.

[0037] The operation unit 12 is configured using existing input devices such as a keyboard, a pointing device (mouse, tablet, etc.), buttons, a touch panel, etc. The operation unit 12 is operated by a user when inputting user instructions to the terminal device 10. The operation unit 12 may be an interface for connecting an input device to the terminal device 10. In this case, the operation unit 12 inputs an input signal generated in the input device in response to a user input to the terminal device 10. The operation unit 12 may be configured using a microphone and a voice recognition device. In this case, the operation unit 12 performs voice recognition on words spoken by the user and inputs character string information of the recognition result to the terminal device 10. In this case, the operation unit 12 may only input voice, and the voice recognition may be performed by the control unit 15. The operation unit 12 may be configured in any way as long as it is capable of inputting user instructions to the terminal device 10.

[0038] The output unit 13 outputs information in a form that can be recognized by the user. The output unit 13 may be, for example, an image display device such as a liquid crystal display or an organic EL (Electro Luminescence) display. The output unit 13 may be an interface for connecting an image display device to the terminal device 10. In this case, the output unit 13 generates a video signal for displaying image data and outputs the video signal to the image display device connected to the output unit 13. The output unit 13 may be a device for outputting sound, such as a speaker. The output unit 13 may be an interface for connecting an audio output device, such as a speaker or headphones, to the terminal device 10. In this case, the output unit 13 generates an audio signal for reproducing audio data and outputs the audio signal to the audio output device connected to the output unit 13.

[0039] The storage unit 14 is configured using a storage device such as a magnetic hard disk drive or a semiconductor storage device. The storage unit 14 stores data used by the control unit 15. The storage unit 14 stores data required when the control unit 15 performs processing.

[0040] The control unit 15 is configured using one or more hardware processors, such as a CPU (Central Processing Unit), and one or more memories (main storage devices). The memory is configured using storage devices, such as RAM (Random Access Memory) and ROM (Read Only Memory). The control unit 15 functions by having one or more hardware processors execute one or more programs stored in the memory to perform various calculations. Note that all or part of the functions of the control unit 15 may be realized using hardware, such as an ASIC (Application Specific Integrated Circuit), a PLD (Programmable Logic Device), or an FPGA (Field Programmable Gate Array). The above programs may be recorded on a computer-readable recording medium. Examples of computer-readable recording media include portable media such as flexible disks, magneto-optical disks, ROMs, CD-ROMs, and semiconductor storage devices (e.g., SSDs: Solid State Drives), as well as storage devices such as hard disks and semiconductor storage devices built into computer systems. The above programs may be transmitted via telecommunications lines.

[0041] The control unit 15 may execute, for example, an application installed on its own device (terminal device 10). A specific example of such an application is an application provided to the terminal device 10 as a dedicated application for the estimation system 100. Another specific example of such an application is a web browser application. Such an application may be installed in the terminal device 10 in advance, or may be downloaded each time an estimation process is executed. For example, when implemented as a web browser application, the application may be configured as follows: When the terminal device 10 connects to a specific web server, the terminal device 10 downloads the application from a device specified by the web server. Then, the terminal device 10 executes the downloaded application. Note that the device specified by the web server may be, for example, the web server itself or another server. The control unit 15 operates according to the program of the application being executed.

[0042] The control unit 15 controls the terminal device 10 in response to user operations and information received from the estimation device 30. For example, the control unit 15 transmits information input by the user operating the operation unit 12 to the estimation device 30 by using the communication unit 11. For example, when information transmitted from the estimation device 30 is received by the communication unit 11 via the network 200, the control unit 15 generates screen data based on the received information and causes the output unit 13 to display the screen data. Such screen data includes images and text transmitted from the estimation device 30. For example, when information transmitted from the estimation device 30 is received by the communication unit 11 via the network 200, the control unit 15 generates voice data based on the received information and causes the output unit 13 to output the voice data.

[0043] Below, a specific example of the operation of the control unit 15 will be described. In the following example, an image display device is used as a specific example of the output unit 13. However, as described above, the output unit 13 does not have to be configured using an image display device, and may be configured using an audio output device, or may be configured using both an image display device and an audio output device.

[0044] The control unit 15 generates screen data having text and images instructing the user to input information required for the estimation device 30 to execute the estimation process. The control unit 15 displays the generated screen data on the output unit 13. Examples of information that the user is instructed to input include additional information about the target scrap. The user inputs the additional information that the user is instructed to input into the terminal device 10 by operating the operation unit 12. For example, when the additional information is input, the control unit 15 may instruct the user to input a command to start the estimation process. The user inputs an instruction to the terminal device 10 to send a command to start the process to the estimation device 30 by operating the operation unit 12. The control unit 15 transmits the input additional information and the instruction to start the estimation process to the estimation device 30 using the communication unit 11.

[0045] The control unit 15 receives the estimation result from the estimation device 30. The control unit 15 generates screen data showing information indicating the estimation result. The control unit 15 displays the generated screen data on the output unit 13. By checking such a display, the user can easily learn information regarding the amount of impurities (copper) contained in the iron scrap that is the subject of estimation (target scrap).

[0046] The imaging device 20 photographs the target scrap to acquire information about the appearance of the target scrap (appearance information). The imaging device 20 may be configured using, for example, a camera (image sensor) that receives visible light from a space containing the target scrap and generates two-dimensional image data. Such a camera is a camera that photographs the target scrap. The imaging device 20 may also be a sensor that acquires information about the three-dimensional shape of the target scrap. In this case, the imaging device 20 may be configured using a device that measures the distance between the imaging device 20 and each point on the surface of the target scrap. More specifically, the imaging device 20 may be configured using a measuring device that measures the distance from the imaging device 20 to each point on the surface of the target scrap by irradiating the target scrap with a laser and measuring the scattered light. A specific example of such an imaging device 20 is LIDAR (Laser Imaging Detection and Ranging). The imaging device 20 transmits the acquired appearance information to the estimation device 30.

[0047] The estimation device 30 estimates the concentration of impurity elements contained in the target scrap. The estimation device 30 is configured using an information processing device such as a personal computer or a server device. FIG. 4 is a schematic block diagram showing a specific example of the functional configuration of the estimation device 30. As shown in the example, the estimation device 30 includes a communication unit 31, a storage unit 32, and a control unit 33.

[0048] The communication unit 31 is a communication device. The communication unit 31 may be configured as, for example, a network interface. The communication unit 31 communicates data with other devices via the network 200 in accordance with the control of the control unit 33. The communication unit 31 may be a device that performs wireless communication or a device that performs wired communication.

[0049] The storage unit 32 is configured using a storage device such as a magnetic hard disk drive or a semiconductor storage device. The storage unit 32 stores data used by the control unit 33. The storage unit 32 may function as an estimation model storage unit 321, for example.

[0050] The estimation model storage unit 321 stores information (hereinafter referred to as "estimation model") used to estimate the concentration of impurity elements contained in the target scrap. The estimation model is generated in advance by a model construction process. Such model construction process may be performed, for example, by another device (e.g., model construction device 40), by the device itself (estimation device 30), or by human labor. The estimation models stored in the estimation model storage unit 321 will be described below.

[0051] The estimation model may be obtained, for example, by a model construction process using known data. The estimation model may be information indicating a trained model obtained by performing supervised learning processing using multiple known data (teacher data) with correct labels, information obtained by performing statistical processing using known data, or other information. Specific examples of learning processing include multivariate analysis, so-called machine learning, deep learning, and the like. As the learning processing, a learning processing for classification or a learning processing for regression may be used depending on the correct labels used in the teacher data. The control unit 33 estimates the concentration of impurity elements contained in the target scrap by using the estimation model stored in the estimation model storage unit 321. The teacher data in this embodiment will be described below.

[0052] The training data is a combination of supplementary information and a correct label (information indicating the concentration of impurity elements) for each collection (hereinafter referred to as a "lot") of a predetermined unit of iron scrap. In this case, the supplementary information corresponds to the explanatory variable, and the information indicating the concentration of impurity elements corresponds to the target variable. A lot may be, for example, the following collections: A collection of iron scrap that was loaded on the same truck when the iron scrap was delivered to the unloading location; A collection of iron scrap delivered from the same scrap dealer within a predetermined time period (e.g., 15 minutes, 30 minutes, 1 hour, etc.); A collection of iron scrap that was loaded on multiple trucks specified by the scrap dealer; A collection of scrap that was attached to a lift magnet by a single application of magnetic force during the unloading process. A lot may be defined in any way as long as it is a collection of iron scrap that is estimated to contain similar amounts of impurities. For example, it may be a collection of iron scrap in an amount that can be transported in one trip by a transport device such as a crane or heavy machinery. In the following description, a collection of iron scraps loaded on the same truck when the iron scraps are transported to the loading and unloading location will be described as one lot.

[0053] The incidental information is information about the iron scrap. The incidental information includes, for example, inspection history information or appearance information. The inspection history information includes attribute information about the iron scrap itself, or information about inspections conducted in advance on the iron scrap (including inspection conditions and inspection results). The data format of the inspection history information may be expressed in any format, such as a string of characters, a number, or an image. The attribute information is, for example, information indicating the source of the iron scrap (a demolished building) or information indicating the time when the iron scrap was generated (when the demolition took place). The information about the inspection is information indicating the supplier of the iron scrap, the iron scrap dismantler, the person who performed the work of removing foreign matter from the iron scrap (sorting work) (hereinafter referred to as the "sorter"), the delivery date and time, the grade of the iron scrap, etc.

[0054] FIG. 5 is a diagram showing a first specific example of training data. The training data shown in FIG. 5 is an example having past inspection information as an explanatory variable and past molten steel composition as a dependent variable. The past inspection information has values ​​of multiple inspection items (x1 to xp) for each lot indicated as No. The past molten steel composition has values ​​of components (y1 to yq) obtained by actually molten steel and measuring it for each lot of the past inspection information. An estimation model may be generated by training an untrained model using these values. FIG. 6 is a diagram showing an example of an estimation model obtained by training using training data such as that shown in FIG. 5. By inputting current inspection information, which is an explanatory variable, into the estimation model, the current predicted molten steel composition, which is the dependent variable, is obtained. The current inspection information indicates the values ​​of each item x1 to xp related to the iron scrap to be estimated. The current predicted molten steel composition indicates the estimated value of each component related to the iron scrap to be estimated.

[0055] FIG. 7 is a diagram showing a second specific example of training data. The training data shown in FIG. 7 is an example having a past image as an explanatory variable and a past molten steel composition as a response variable. The past image is one or more images captured at a predetermined process for each lot indicated by No. In the example of FIG. 7, one image is used for each lot. The past molten steel composition has values ​​of components (y1 to yq) obtained by actually molten steel and measuring each lot of the past image. An estimation model may be generated by training an untrained model using these values. FIG. 8 is a diagram showing an example of an estimation model obtained by training using training data such as that shown in FIG. 7. By inputting the current image (No. A1), which is an explanatory variable, into the estimation model, the current predicted molten steel composition, which is the response variable, is obtained. The current image indicates an image captured at a predetermined process for the iron scrap to be estimated. The current predicted molten steel composition indicates an estimated value of each component for the iron scrap to be estimated.

[0056] FIG. 9 is a diagram showing a third specific example of training data. The training data shown in FIG. 9 is an example having past inspection information and past images as explanatory variables and having past molten steel composition as a dependent variable. The past inspection information and past images are as described above. An estimation model may be generated by training an untrained model using these values. FIG. 10 is a diagram showing an example of an estimation model obtained by training using training data such as that shown in FIG. 9. By inputting the current inspection information and current image (e.g., a pre-melting scrap image), which are explanatory variables, into the estimation model, the current predicted molten steel composition, which is the dependent variable, can be obtained.

[0057] Other information such as the type and number of foreign objects estimated to be mixed in the steel scrap, the grade of the scrap determined by a grade determination system, and composition data of all parts that make up the scrap may also be included as inspection performance information. The composition data may be, for example, the content ratio of each type of part, such as thin plate, reinforcing bar, or H-beam.

[0058] FIG. 11 is a diagram showing a fourth specific example of training data. The training data shown in FIG. 11 is an example having past inspection information and past contraindicated substance detection results as explanatory variables and past molten steel composition as a dependent variable. The past inspection information is as described above. The past contraindicated substance detection results include values ​​for multiple contraindicated substances (tramp element components z1-zq) for each lot indicated as No. The past molten steel composition includes values ​​for components (y1-yq) obtained by actually molten steel and measuring each lot of the past inspection information. An estimation model may be generated by training an untrained model using these values. FIG. 12 is a diagram showing an example of an estimation model obtained by training using training data such as that shown in FIG. 11. The current inspection information and current contraindicated substance detection results, which are explanatory variables, are input to the estimation model to obtain the current predicted molten steel composition, which is the dependent variable. The current inspection information indicates the values ​​of each item x1-xp related to the steel scrap to be estimated. The current contraindicated substance detection results indicate the values ​​of each item z1-zq related to the contraindicated steel in the steel scrap to be estimated. The presently predicted molten steel composition indicates the estimated value of each component of the iron scrap being estimated.

[0059] Furthermore, the training data is not limited to the above examples, and various additional information may be used. Other examples of training data include information indicating the cause of iron scrap generation and information indicating suppliers. Examples of causes of generation include the demolition of buildings and the generation of scrap during steel processing.

[0060] In addition, prohibited materials in steel scrap include impurities such as so-called tramp elements, sealed objects (such as gas cylinders) that may explode when melted, and iron with a large amount of non-conductive materials (such as insulation). Among these prohibited materials, those containing tramp elements are relatively common. Therefore, the concentration of impurity elements in steel scrap correlates with the amount of prohibited materials contained in the steel scrap. Some of the prohibited materials contained in steel scrap are removed by sorters at suppliers before the steel scrap is delivered. Therefore, the amount of prohibited materials contained in steel scrap correlates with the supplier. For example, among multiple suppliers, there are some that remove more prohibited materials and others that remove less. The amount of prohibited materials contained in steel scrap also correlates with the sorter. For example, among multiple sorters belonging to the same supplier, there are some that remove more prohibited materials and others that remove less.

[0061] The amount of prohibited substances contained in steel scrap varies greatly depending on the type of structure from which the steel scrap was obtained. Therefore, the amount of prohibited substances contained in steel scrap correlates with the structure (origin) from which the steel scrap was obtained. Furthermore, steel scrap obtained from the same structure is delivered in large quantities around the time the structure is demolished. Therefore, the amount of prohibited substances contained in steel scrap correlates with the time when the steel scrap was delivered and the time when the steel scrap was generated.

[0062] The type and amount of impurities contained in iron scrap can be estimated based on shape and color information obtained from its appearance. For example, it is possible to determine the grade of heavy scrap based on shape (especially thickness). The amount of rust and the color of coated copper wire can also be determined based on color. Furthermore, the presence of prohibited objects such as motors and electrical distribution boards can be determined based on information obtained from their appearance. Therefore, the concentration of impurity elements contained in iron scrap correlates with information obtained from its appearance (such as two-dimensional images).

[0063] The appearance information is information acquired by the imaging device as described above. In the following description, an example will be described in which a two-dimensional image acquired by receiving visible light is used as the appearance information.

[0064] In generating the estimation model, both the inspection performance information and the two-dimensional image (appearance information) of the incidental information may be used, or only one of them may be used. If only the inspection performance information is used in generating the estimation model, only the inspection performance information is also used in the estimation process using the estimation model. If only the two-dimensional image is used in generating the estimation model, only the two-dimensional image is also used in the estimation process using the estimation model. If the inspection performance information and the two-dimensional image are used in generating the estimation model, it is desirable that the inspection performance information and the two-dimensional image are used in the estimation process using the estimation model, but only one of the inspection performance information and the two-dimensional image may be used. Furthermore, the information used as the appearance information may be any information that can indicate the appearance of the iron scrap. For example, information indicating the three-dimensional shape acquired using a three-dimensional shape measurement system such as LIDAR may be used as the appearance information. Furthermore, other types of image information, such as multispectral images and infrared images, may be used as the appearance information.

[0065] When only inspection performance information is used to generate the estimation model, a regression model such as multiple regression or a decision tree may be used as the estimation model. Alternatively, a model that outputs a predicted value as a probability distribution, such as Gaussian process regression, may be used. Furthermore, a classification model such as a support vector machine (SVM) that outputs which class among multiple classes indicating a range of predicted values ​​belongs to may be used as the estimation model. When only two-dimensional images are used to generate the estimation model, a neural network such as a convolutional neural network (CNN) or other learning algorithms may be used as the estimation model. Even when only two-dimensional images are used to generate the estimation model, a regression model that outputs a predicted value may be used, or a classification model that outputs which class among multiple classes indicating a range of predicted values ​​belongs to may be used.

[0066] When inspection performance information and two-dimensional images are used to generate an estimation model, a multimodal learning model (e.g., multimodal deep learning) may be used for the estimation model. Specific examples of image-based multimodalization techniques (meta-information input techniques) include early-fusion and late-fusion. Early-fusion converts inspection performance and image information into vectors of the same dimension so that a single prediction model can be used. As a result, inference results can be obtained from a single prediction model. Late-fusion uses a TE component prediction model (e.g., Gaussian process regression) using inspection performance and a TE component prediction model (e.g., CNN) using image information independently, and weights the respective prediction results to obtain a final prediction result. Multimodalization techniques are not limited to these specific examples, and other techniques may also be applied. When inspection performance information and two-dimensional images are used to generate an estimation model, a regression model that outputs a predicted value may be used, or a classification model that outputs which class the predicted value belongs to among multiple classes indicating a range of the predicted value may be used.

[0067] Next, the control unit 33 will be described. The control unit 33 is configured using one or more hardware processors such as a CPU and one or more memories (main storage devices). The memory is configured using storage devices such as RAM and ROM (Read Only Memory). The control unit 33 functions by having one or more hardware processors execute one or more programs stored in the memory to perform various calculations. The control unit 33 functions as, for example, an information control unit 331, an estimation unit 332, and an output control unit 333. Note that all or part of the functions of the control unit 33 may be realized using hardware such as an ASIC, PLD, or FPGA. The above programs may be recorded on a computer-readable recording medium. Examples of computer-readable recording media include portable media such as flexible disks, magneto-optical disks, ROMs, CD-ROMs, and semiconductor storage devices (e.g., SSDs), as well as storage devices such as hard disks and semiconductor storage devices built into a computer system. The above programs may be transmitted via a telecommunications line.

[0068] The information control unit 331 acquires information from other devices such as the terminal device 10 and the imaging device 20. A specific example of the acquired information is, for example, supplementary information on the target scrap. Specifically, the information control unit 331 may acquire either or both of inspection performance information and two-dimensional images (appearance information). The information control unit 331 records the acquired information in a storage device such as a memory. The information control unit 331 transmits information to other devices such as the terminal device 10. A specific example of the information to be transmitted is, for example, information estimated by the estimation unit 332. Such exchange of information between the information control unit 331 and other devices may be performed by, for example, communication via the communication unit 31.

[0069] Next, the estimation unit 332 will be described. The estimation unit 332 estimates the concentration of impurity elements contained in the target scrap based on the incidental information acquired by the information control unit 331 and the estimation model stored in the estimation model storage unit 321. A specific example of such processing will be described. The estimation unit 332 acquires incidental information corresponding to the estimation model. In other words, the estimation unit 332 acquires information equivalent to the explanatory variables of the known data used in acquiring the estimation model. The estimation unit 332 estimates the concentration of impurity elements contained in the target scrap using information equivalent to the acquired explanatory variables (information included in the incidental information) and the estimation model.

[0070] The output control unit 333 outputs information indicating the estimation result by the estimation unit 332. The output control unit 333 may output the information indicating the estimation result by, for example, transmitting it to the terminal device 10. The output control unit 333 may output the information indicating the estimation result by controlling an output device connected to the device itself. For example, the output control unit 333 may cause an image display device connected to the device itself to display a character string or an image indicating the estimation result. For example, the output control unit 333 may cause a speaker connected to the device itself to output a sound indicating the estimation result.

[0071] 13 is a flowchart showing a specific example of processing by the estimation device 30. First, the information control unit 331 acquires incidental information about the target scrap from another device (the terminal device 10 or the imaging device 20) (step S101). The estimation unit 332 estimates the concentration of impurity elements contained in the target scrap using the acquired incidental information and an estimation model (step S102). The estimated concentration may be calculated, for example, using the following formula: mass percent concentration (%) = (mass (g) of TE component / mass (g) of scrap). Then, the output control unit 333 outputs the estimation result (step S103).

[0072] 14 is a schematic block diagram showing a specific example of the functional configuration of the model construction device 40. The model construction device 40 is configured using an information processing device such as a personal computer or a server device. The model construction device 40 includes a communication unit 41, a storage unit 42, and a control unit 43.

[0073] The communication unit 41 is a communication device. The communication unit 41 may be configured as, for example, a network interface. The communication unit 41 communicates data with other devices via the network 200 in accordance with the control of the control unit 43. The communication unit 41 may be a device that performs wireless communication or a device that performs wired communication.

[0074] The storage unit 42 is configured using a storage device such as a magnetic hard disk drive or a semiconductor storage device. The storage unit 42 stores data used by the control unit 43. The storage unit 42 may function as, for example, a known data storage unit 421 and an estimation model storage unit 422.

[0075] The known data storage unit 421 stores known data used in the model construction process executed in the model construction device 40. The known data may be, for example, teacher data used in supervised learning processes. The known data stored in the known data storage unit 421 is, for example, data including incidental information and a correct answer label related to the concentration of impurity elements contained in the target scrap having the incidental information. In this case, the incidental information included in the known data corresponds to the explanatory variable, and the concentration of impurity elements contained in the target scrap corresponds to the objective variable. The incidental information included in the known data may be, for example, a two-dimensional image (appearance information) acquired by the imaging device 20.

[0076] The correct labels included in the known data can be obtained by analyzing the iron scrap for which additional information has been obtained as known data. Specifically, they can be obtained, for example, by the following analysis. For example, the amount of impurities (e.g., copper) contained in the iron scrap may be obtained by performing optical analysis, such as laser, X-ray, or neutron beam analysis, on the entire iron scrap. For example, when multiple iron scraps are mixed and melted, the correct labels for each known data item may be calculated from the results of measuring the concentration of impurity elements in the molten steel using multiple blend melting records in which the blending ratios of the mixed iron scraps are changed while keeping the same blend combination. For example, the correct labels may be obtained by melting the scrap lot for which the TE concentration is desired alone and measuring the components of the molten steel.

[0077] The estimation model storage unit 422 stores an estimation model obtained by a model construction process using the known data stored in the known data storage unit 421 .

[0078] The control unit 43 is configured using a processor such as a CPU and a memory. The control unit 43 functions as an information control unit 431 and a model construction control unit 432 when the processor executes a program. All or part of the functions of the control unit 43 may be realized using hardware such as an ASIC, PLD, or FPGA. The above program may be recorded on a computer-readable recording medium. Examples of computer-readable recording media include portable media such as flexible disks, magneto-optical disks, ROMs, CD-ROMs, and semiconductor storage devices (e.g., SSDs), as well as storage devices such as hard disks and semiconductor storage devices built into computer systems. The above program may be transmitted via a telecommunications line.

[0079] The information control unit 431 controls the input and output of information. For example, the information control unit 431 acquires known data from other devices (information processing devices or storage media) and records it in the known data storage unit 421. For example, the information control unit 431 transmits the estimation model stored in the estimation model storage unit 422 to other devices (e.g., the estimation device 30).

[0080] The model construction control unit 432 constructs an estimation model by performing a model construction process using the known data stored in the known data storage unit 421. The model construction control unit 432 may construct an estimation model, for example, by performing a supervised learning process using the supplementary information stored in the known data storage unit 421 as training data. Note that the process performed by the model construction control unit 432 does not need to be limited to a supervised learning process. For example, the model construction control unit 432 may construct an estimation model by performing a learning process using a statistical method such as multivariate analysis using the known data.

[0081] As a specific example of such a learning process, for example, a regression machine learning process may be used, or other learning techniques such as neural networks and deep learning may be used. The model construction control unit 432 records the generated estimation model in the estimation model storage unit 422. The estimation model obtained by the model construction control unit 432 may be transmitted to the estimation device 30 and recorded in the estimation model storage unit 321 of the estimation device 30. By providing explanatory variables including auxiliary information as input, such an estimation model can obtain, as output, an estimated value of the concentration of an impurity element according to the estimation model.

[0082] 15 is a flowchart showing a specific example of processing by the model construction device 40. First, the information control unit 431 acquires known data (step S201). The known data may be input by a user, acquired via communication from another information device, or acquired from a recording medium connected to the model construction device 40, for example. The model construction control unit 432 executes a model construction process using the known data and records an estimated model in the estimated model storage unit 422 (step S202).

[0083] In the estimation system 100 configured in this manner, the concentration of impurity elements contained in the target scrap is estimated based on an estimation model obtained based on known data of other iron scrap and on the additional information of the target scrap. This makes it possible to estimate the concentration of impurity elements in the target scrap more accurately than methods that use images and additional information individually.

[0084] Furthermore, estimation system 100 estimates the concentrations of impurity elements contained in target scrap based on information obtained without melting the target scrap. Therefore, estimation system 100 makes it possible to estimate the concentrations of impurity elements contained in the target scrap before actually melting the target scrap. Therefore, it becomes possible to appropriately blend multiple target scraps based on the concentrations of impurity elements, and it is possible to prevent the concentration values ​​of impurity elements after melting from falling outside the specification range.

[0085] Furthermore, by making predictions before the target scrap is actually melted using the estimation system 100, it becomes possible to prevent contamination with copper and other substances. Furthermore, since predictions can be made before the scrap is melted, when TE components exceeding the standard are measured, it becomes possible to take measures such as adding a process to remove taboo substances that could be the cause.

[0086] (Variation) In this embodiment, the terminal device 10 and the estimation device 30 are configured as separate devices, but they may also be configured as an integrated device. Fig. 16 is a diagram showing a modified example of the estimation device 30 configured in this manner. The estimation device 30 shown in Fig. 16 includes an operation unit 34 and an output unit 35. The operation unit 34 and the output unit 35 of the estimation device 30 shown in Fig. 16 function in the same manner as the operation unit 12 and the output unit 13 of the terminal device 10, respectively. The control unit 33 operates in response to an operation on the operation unit 34, and outputs information using the output unit 35.

[0087] In this embodiment, the estimation device 30 and the model construction device 40 are configured as separate devices, but they may also be configured as an integrated device. FIG. 17 is a diagram showing a modified example of the estimation device 30 configured in this manner. The storage unit 32 of the estimation device 30 shown in FIG. 17 also functions as a known data storage unit 322. The control unit 33 of the estimation device 30 shown in FIG. 17 also functions as a model construction control unit 334. The known data storage unit 322 functions in the same way as the known data storage unit 421 of the model construction device 40. The model construction control unit 334 functions in the same way as the model construction control unit 432 of the model construction device 40.

[0088] The estimation device 30 may be implemented using a plurality of information processing devices. For example, the estimation device 30 may be implemented using a device such as a cloud. For example, in the estimation device 30, the storage unit 32 and the control unit 33 may be implemented in different information processing devices. For example, the storage unit 32 of the estimation device 30 may be distributed and implemented across a plurality of information processing devices. The model construction device 40 may be implemented using a plurality of information processing devices. For example, the model construction device 40 may be implemented using a device such as a cloud. For example, in the model construction device 40, the storage unit 42 and the control unit 43 may be implemented in different information processing devices. For example, the storage unit 42 of the model construction device 40 may be distributed and implemented across a plurality of information processing devices.

[0089] The estimation model stored in the estimation model storage unit 321 of the estimation device 30 may be updated at a predetermined timing. For example, the estimation model storage unit 321 may be configured to store an estimation model generated based only on known data obtained for iron scrap generated within a predetermined period going back in time. This configuration allows the estimation process to be performed using an estimation model that reflects particularly the most recent incidental information. As a result, the accuracy of the estimation process can be improved. Another specific example of the predetermined timing is the timing when a significant change in data trend is detected to satisfy a predetermined criterion. More specifically, the timing may be the timing when a change that satisfies a predetermined criterion is detected by unsupervised learning, or another algorithm for detecting a trend change may be used.

[0090] A spectral sensor may be used instead of the imaging device 20. In this case, spectral information of light received from the target scrap is obtained as appearance information. In this case, an estimation model obtained by performing a model construction process using known data containing similar spectral information is used in the estimation process. The estimation device 30 performs the estimation process using the spectral information obtained from the spectral sensor. The estimation device 30 may perform the estimation process using the spectral information without using any additional information, or may perform the estimation process using the additional information and the spectral information. The estimation device 30 may perform the estimation process using the additional information, the spectral information, and image data. In either case, the model construction process is performed using known information including each piece of information used. This configuration makes it possible to obtain information that is difficult to detect using information obtained with visible light or three-dimensional shape information (e.g., information regarding the material of components contained in the iron scrap) and perform the estimation process with higher accuracy.

[0091] [Second embodiment] FIG. 18 is a schematic block diagram showing the system configuration of a second embodiment of the estimation system 100 of the present invention. The estimation system 100 of the second embodiment further includes a taboo object estimation device 50. The taboo object estimation device 50 estimates information about taboo objects contained in the target scrap. Specific examples of the information about taboo objects include the types of taboo objects contained in the target scrap, the number of each type of taboo object contained in the target scrap, and the total number of taboo objects contained in the target scrap. The information about taboo objects may include all of the above information or a portion of it. In the following description, an example will be described in which the information about taboo objects includes the types of taboo objects contained in the target scrap and the number of each type of taboo object contained in the target scrap. The estimation device 30 performs estimation processing using the estimation results from the taboo object estimation device 50. The second embodiment of the estimation system 100 will be described below.

[0092] The taboo object estimation device 50 estimates taboo objects contained in the target scrap. The taboo object estimation device 50 is configured using an information processing device such as a personal computer or a server device. FIG. 19 is a schematic block diagram showing a specific example of the functional configuration of the taboo object estimation device 50. As shown in the example, the taboo object estimation device 50 includes a communication unit 51, a memory unit 52, and a control unit 53.

[0093] The communication unit 51 is a communication device. The communication unit 51 may be configured as, for example, a network interface. The communication unit 51 communicates data with other devices via the network 200 in accordance with the control of the control unit 53. The communication unit 51 may be a device that performs wireless communication or a device that performs wired communication.

[0094] The storage unit 52 is configured using a storage device such as a magnetic hard disk drive or a semiconductor storage device. The storage unit 52 stores data used by the control unit 53. The storage unit 52 may function as a taboo object estimation model storage unit 521, for example.

[0095] The taboo substance estimation model storage unit 521 stores information used to estimate taboo substances contained in the target scrap (hereinafter referred to as the "taboo substance estimation model"). The taboo substance estimation model is generated in advance by a model construction process. Such a model construction process may be executed, for example, by another device (e.g., the model construction device 40), by the device itself (the taboo substance estimation device 50), or by human labor. The taboo substance estimation model stored in the taboo substance estimation model storage unit 521 will be described below.

[0096] The taboo substance estimation model may be obtained, for example, by a model construction process using known data. The taboo substance estimation model may be information indicating a trained model obtained by performing supervised learning using a plurality of known data (teacher data) having correct labels, information obtained by performing statistical processing using known data, or other information. Specific examples of learning processes include multivariate analysis, so-called machine learning, deep learning, and the like. As the learning process, a learning process for classification or a learning process for regression may be used depending on the correct labels used in the teacher data. The control unit 53 estimates the taboo substances contained in the target scrap by using the estimation model stored in the taboo substance estimation model storage unit 521. The teacher data in this embodiment will be described below.

[0097] The training data is a combination of an image of a taboo object and a correct answer label (information indicating the type of taboo object contained in the image). Taboo objects in iron scrap are objects that contain a large amount of impurities, such as so-called tramp elements. As the image of the taboo object, an image of the taboo object alone may be used, or an image of the taboo object mixed in the iron scrap may be used. Alternatively, an image of the taboo object mixed in the iron scrap may be generated by superimposing an image of the taboo object alone on an image of the iron scrap, and the generated image may be used as the image of the taboo object. In these cases, a portion (partial image) of the image of the target scrap may be input to the taboo object estimation model, and an estimation result indicating whether the partial image contains a taboo object may be output.

[0098] The training data may be a combination of an image of the target scrap and a correct answer label (information indicating the type of prohibited substance contained in the image). For example, an image of a collection of multiple iron scraps may be used, such as an image captured by the imaging device 20. In this case, the correct answer label may include the number of each type of prohibited substance contained in the target scrap shown in the image.

[0099] Next, the control unit 53 will be described. The control unit 53 is configured using one or more hardware processors such as a CPU and one or more memories (main storage devices). The memory is configured using storage devices such as RAM and ROM (Read Only Memory). The control unit 53 functions by having one or more hardware processors execute one or more programs stored in the memory to perform various calculations. The control unit 53 functions as, for example, an information control unit 531, a taboo object estimation unit 532, and an output control unit 533. Note that all or part of the functions of the control unit 53 may be realized using hardware such as an ASIC, PLD, or FPGA. The above programs may be recorded on a computer-readable recording medium. Examples of computer-readable recording media include portable media such as flexible disks, magneto-optical disks, ROMs, CD-ROMs, and semiconductor storage devices (e.g., SSDs), as well as storage devices such as hard disks and semiconductor storage devices built into a computer system. The above programs may be transmitted via a telecommunications line.

[0100] The information control unit 531 acquires information from other devices such as the terminal device 10 and the imaging device 20. A specific example of the acquired information is, for example, an image captured by the imaging device 20. The information control unit 531 records the acquired information in a storage device such as a memory. The information control unit 531 transmits information to other devices such as the terminal device 10 and the estimation device 30. A specific example of the transmitted information is, for example, information estimated by the taboo object estimation unit 532. Such exchange of information between the information control unit 531 and other devices may be performed by communication using the communication unit 51, for example.

[0101] Next, the taboo object estimation unit 532 will be described. The taboo object estimation unit 532 estimates taboo objects contained in the target scrap based on the image acquired by the information control unit 531 and the taboo object estimation model stored in the taboo object estimation model storage unit 521. A specific example of such processing will be described. The taboo object estimation unit 532 acquires an image corresponding to the taboo object estimation model. In other words, the taboo object estimation unit 532 acquires information equivalent to the explanatory variables of the known data used in acquiring the taboo object estimation model. The taboo object estimation unit 532 estimates taboo objects contained in the target scrap using information equivalent to the acquired explanatory variables (images of the same size) and the taboo object estimation model.

[0102] The output control unit 533 outputs information indicating the estimation result by the taboo object estimation unit 532. The output control unit 533 outputs the information by transmitting the information indicating the estimation result to the estimation device 30. The output control unit 533 may further transmit the information indicating the estimation result to a device such as the terminal device 10. The output control unit 533 may output the information indicating the estimation result by controlling an output device connected to the device itself. For example, the output control unit 533 may display a character string or an image indicating the estimation result on an image display device connected to the device itself. For example, the output control unit 533 may output a sound indicating the estimation result from a speaker connected to the device itself.

[0103] 20 is a flowchart showing a specific example of processing by the proscribed object estimation device 50. First, the information control unit 531 acquires an image of the target scrap from another device (imaging device 20) (step S301). The proscribed object estimation unit 532 estimates the proscribed objects contained in the target scrap using the acquired image and a proscribed object estimation model (step S302). Then, the output control unit 533 outputs the estimation result to the estimation device 30 (step S303).

[0104] Unlike the first embodiment, the estimation unit 332 of the estimation device 30 shown in FIG. 18 in the second embodiment uses information about prohibited substances contained in the target scrap as one of the additional information. Specifically, the estimation unit 332 performs an estimation process for the impurity element concentration of the target scrap using information indicating the type and number of prohibited substances contained in the target scrap as explanatory variables. The information used in the estimation process is the estimation results of the prohibited substance estimation device 50. That is, the estimation unit 332 receives the estimation results from the prohibited substance estimation device 50, and the estimation model in the second embodiment is generated by performing a model construction process using known data whose additional information includes information about prohibited substances. Therefore, the model construction device 40 in the second embodiment stores known data whose additional information includes information about prohibited substances in the known data storage unit 421, and the model construction control unit 432 generates the estimation model using this known data.

[0105] In the estimation system 100 configured in this manner, when estimating the concentrations of impurity elements contained in the target scrap, information about the taboo substances estimated to be contained in the target scrap (e.g., the type and number of such substances) is used as an explanatory variable. Because taboo substances contain many impurities, it can be said that information about the taboo substances contained in the target scrap is highly correlated with the concentrations of impurity elements contained in the target scrap. Therefore, by configuring in this manner, it becomes possible to more accurately estimate the concentrations of impurity elements in the target scrap.

[0106] (Variation) In this embodiment, the estimation device 30 and the contraindicated object estimation device 50 are configured as separate devices, but they may also be configured as an integrated device. FIG. 21 is a diagram showing a modified example of the estimation device 30 configured in this manner. The memory unit 32 of the estimation device 30 shown in FIG. 21 also functions as a contraindicated object estimation model memory unit 323. The control unit 33 of the estimation device 30 shown in FIG. 21 also functions as a contraindicated object estimation unit 335. The contraindicated object estimation model memory unit 323 functions in the same way as the contraindicated object estimation model memory unit 521 of the contraindicated object estimation device 50. The contraindicated object estimation unit 335 functions in the same way as the contraindicated object estimation unit 532 of the contraindicated object estimation device 50.

[0107] The contraindicated object estimation device 50 may be implemented using a plurality of information processing devices. For example, the contraindicated object estimation device 50 may be implemented using a device such as a cloud. For example, in the contraindicated object estimation device 50, the storage unit 52 and the control unit 53 may be implemented in different information processing devices. For example, the storage unit 52 of the contraindicated object estimation device 50 may be implemented in a distributed manner in a plurality of information processing devices.

[0108] [Third embodiment] FIG. 22 is a schematic block diagram showing the system configuration of a third embodiment of the estimation system 100 of the present invention. The estimation system 100 of the third embodiment further includes a removal target determination device 60 and a weight measurement device 400 in addition to the estimation system 100 of the second embodiment. The removal target determination device 60 is configured to set an upper limit for the concentration of impurity elements that can be tolerated in the target scrap (hereinafter referred to as the "impurity upper limit value"). The impurity upper limit value may be determined as a concentration or as an absolute amount (mass) per weight of the scrap being handled. The removal target determination device 60 determines which of the taboo substances contained in the target scrap should be removed (hereinafter referred to as the "removal target taboo substances") so that the concentration of the impurity elements in the target scrap is below the impurity upper limit value. By having an operator remove the taboo substances from the target scrap according to the determination result of the removal target determination device 60, the concentration of the impurity elements in the target scrap can be adjusted to be below the impurity upper limit value. The third embodiment of the estimation system 100 will now be described.

[0109] The weight measuring device 400 measures the weight of the target scrap in the estimation system 100. The weight measuring device 400 may, for example, measure the weight of the target scrap itself. In this case, the weight of the target scrap itself may be measured, for example, by placing the target scrap on a weighing scale, as described below. The weight measuring device 400 may, for example, measure the weight of the container containing the target scrap and the target scrap together. In this case, the weight measuring device 400 may also measure the weight of the container after all the target scrap has been removed. The weight of the target scrap may be calculated by subtracting the measurement result of the weight of the container after all the target scrap has been removed from the measurement result of the weight of the container and the target scrap. This calculation may be performed by the weight measuring device 400 or by a device such as the estimation device 30 or the removal target determination device 60. The weight measuring device 400 may, for example, measure the weight of the mobile object (e.g., a truck vehicle) containing the target scrap and the target scrap together. In this case, the weight measuring device 400 may also measure the weight of the moving body after all of the target scrap has been removed. The weight of the target scrap may be calculated by subtracting the measurement result of the weight of the moving body after all of the target scrap has been removed from the measurement result of the weight of the moving body and the target scrap. This calculation may be performed by the weight measuring device 400, or by a device such as the estimation device 30 or the removal target determination device 60.

[0110] The weight measuring device 400 may be configured, for example, using a weighing scale having a device on top of which the target scrap is loaded and which weighs the object loaded on the device. The weight measuring device 400 may be configured, for example, using a vehicle weighing scale that is embedded in the ground and measures the weight of a vehicle positioned above it. Other configurations may also be adopted for the weight measuring device 400. The weight measuring device 400 transmits information indicating the measurement results to the estimation device 30 and the removal target determination device 60.

[0111] The removal target determination device 60 determines which of the prohibited substances contained in the target scrap are to be removed so that the impurity content is below the upper limit value. The removal target determination device 60 is configured using an information processing device such as a personal computer or a server device. FIG. 23 is a schematic block diagram showing a specific example of the functional configuration of the removal target determination device 60. As shown in the example, the removal target determination device 60 includes a communication unit 61, a memory unit 62, and a control unit 63.

[0112] The communication unit 61 is a communication device. The communication unit 61 may be configured as, for example, a network interface. The communication unit 61 communicates data with other devices via the network 200 in accordance with the control of the control unit 63. The communication unit 61 may be a device that performs wireless communication or a device that performs wired communication.

[0113] The storage unit 62 is configured using a storage device such as a magnetic hard disk drive or a semiconductor storage device. The storage unit 62 stores data used by the control unit 63. The storage unit 62 may function as a taboo object information storage unit 621, for example.

[0114] The taboo substance information storage unit 621 stores taboo substance information, which is information related to the weight of taboo substances. FIG. 24 is a diagram showing a specific example of taboo substance information. The taboo substance information shown in FIG. 24 has three values: taboo substance name, copper weight, and average weight. The taboo substance name indicates the name of the type of each taboo substance. The copper weight indicates the average weight of the impurity (copper) contained in each type of taboo substance. Note that in this embodiment, copper is used as an example of an impurity contained in a taboo substance, but the weight of other impurities may also be registered in the taboo substance information. The average weight indicates the average weight of each type of taboo substance. For example, it is empirically known that copper accounts for approximately 20% of the total weight of a motor. It is assumed that other taboo substances are organized in a similar manner.

[0115] Next, the control unit 63 will be described. The control unit 63 is configured using one or more hardware processors such as a CPU and one or more memories (main storage devices). The memory is configured using storage devices such as RAM and ROM (Read Only Memory). The control unit 63 functions by having one or more hardware processors execute one or more programs stored in the memory to perform various calculations. The control unit 63 functions as, for example, an information control unit 631, a removal target determination unit 632, and an output control unit 633. Note that all or part of the functions of the control unit 63 may be realized using hardware such as an ASIC, PLD, or FPGA. The above programs may be recorded on a computer-readable recording medium. Examples of computer-readable recording media include portable media such as flexible disks, magneto-optical disks, ROMs, CD-ROMs, and semiconductor storage devices (e.g., SSDs), as well as storage devices such as hard disks and semiconductor storage devices built into a computer system. The above programs may be transmitted via telecommunications lines.

[0116] The information control unit 631 acquires information from other devices such as the terminal device 10, the estimation device 30, and the contraindicated substance estimation device 50. A specific example of the acquired information is, for example, the estimation result of the contraindicated substance estimation device 50. Another specific example of the acquired information is, for example, information on the upper limit value of impurities. The information control unit 631 records the acquired information in a storage device such as a memory. The information control unit 631 transmits information to other devices such as the terminal device 10 and the estimation device 30. A specific example of the transmitted information is, for example, information determined by the removal target determination unit 632 (information indicating the contraindicated substance to be removed). Such exchange of information between the information control unit 631 and other devices may be performed by, for example, communication using the communication unit 61.

[0117] The removal target determination unit 632 determines the type and amount of the impurity-targeted prohibited substance from the target scrap based on the weight of the target scrap, the estimated concentration of the impurity elements contained in the target scrap, information about the prohibited substances contained in the target scrap, and the impurity upper limit value. That is, the removal target determination unit 632 receives as input the estimated concentration of the impurity elements contained in the target scrap, information about the prohibited substances contained in the target scrap, and the impurity upper limit value. The removal target determination unit 632 performs a predetermined process (removal necessity determination process) to determine whether or not the prohibited substances need to be removed so that the impurity concentration falls below the impurity upper limit value, and outputs the determination result. For example, if this process is performed when unloading from the truck bed using the transport device 300, the process may be performed to reduce the impurity element concentration below the impurity upper limit value for each lot lifted by the lift magnet 301 for unloading.

[0118] When this process is performed when unloading scrap from a truck bed using the transport device 300, the process may be performed as follows. Each time scrap is lifted from the truck bed by the lift magnet 301 for unloading, an image of the target scrap loaded on the truck bed is acquired. Then, for each newly acquired image, processing is performed so that the impurity element concentration falls below the upper impurity limit. Each time scrap is lifted by the lift magnet 301, a new surface layer of the target scrap loaded on the truck bed is exposed and captured in the image. This makes it possible to properly identify target scrap loaded inside or at the bottom, which was not captured in the image when the truck first arrived. The environment in which this configuration is adopted does not need to be limited to an environment in which the target scrap is lifted from the truck bed. For example, it can be applied to an environment in which a large amount of target scrap is piled up in a predetermined area, such as the ground, and then gradually transported to another area using a transport device such as the lift magnet 301.

[0119] A specific example of the removal necessity determination process will be described. The removal target determination unit 632 first obtains an estimation result of information regarding the concentration of impurity elements contained in the target scrap to be determined. Such estimation results may be obtained, for example, by performing estimation processing using an estimation model stored in the estimation model storage unit 321. The removal target determination unit 632 may determine whether removal is necessary based on whether the estimation result (estimated value) of the concentration of impurity elements contained in the target scrap is less than a predetermined threshold. For example, if the estimated value is less than the predetermined threshold, it determines that removal is unnecessary, and if the estimated value is equal to or greater than the predetermined threshold, it determines that removal is necessary. The predetermined threshold may be set for each type of impurity element.

[0120] 25 is a flowchart showing a specific example of the processing of the removal target determination unit 632. A specific example of the processing of the removal target determination unit 632 will be described below. The removal target determination unit 632 acquires an estimated value of the concentration of impurity elements contained in the target scrap (step S401). The removal target determination unit 632 determines whether the estimated value of the concentration of the impurity elements is less than the impurity upper limit value (step S402). If the estimated value of the concentration of the impurity elements exceeds the impurity upper limit value (step S402-NO), the removal target determination unit 632 determines taboo substances that should be removed (taboo substances to be removed) so that the concentration of the impurity elements falls below the impurity upper limit value (step S403).

[0121] In the processing of step S403, for example, the removal target determination unit 632 determines the taboo substances to be removed so that the concentration of impurity elements in the target scrap after removing the taboo substances to be removed is less than the upper limit value. For example, by referencing the taboo substance information stored in the taboo substance information storage unit 621, the average value of the impurity (copper) contained in each type of taboo substance and the average weight of the taboo substance itself can be determined. If the impurity upper limit value is given as a percentage, the total amount of impurities to be removed can be calculated by multiplying the weight of the target scrap by the value obtained by subtracting the impurity upper limit value from the estimated concentration of the impurity element. The removal target determination unit 632 determines the taboo substances to be removed so that the total amount of impurities contained in the taboo substances to be actually removed exceeds the total amount calculated here.

[0122] At this time, if the estimation result of the prohibited substance estimation device 50 has been obtained, the removal target determination unit 632 may select the prohibited substances to be removed from the types and number of prohibited substances included in the estimation result of the prohibited substance estimation device 50. By having the removal target determination unit 632 determine the prohibited substances to be removed in this way, it becomes possible to output information to the worker about prohibited substances that can be more realistically removed, and it becomes easier to reduce the concentration of impurity elements contained in the target scrap to below the impurity upper limit value.

[0123] The output control unit 633 outputs information indicating the determination result by the removal target determination unit 632. The output control unit 633 may output the information indicating the determination result by, for example, transmitting it to the terminal device 10. The output control unit 633 may output the information indicating the determination result by controlling an output device connected to the terminal device. For example, the output control unit 633 may cause an image display device connected to the terminal device to display a character string or an image indicating the determination result. For example, the output control unit 633 may cause a speaker connected to the terminal device to output a sound indicating the determination result.

[0124] FIG. 26 is a diagram showing an application example of the estimation system 100 of this embodiment. In the application example shown in FIG. 26, a weight measuring device 400 is further provided in addition to the application example shown in FIG. 2. In FIG. 26, for example, the weight measuring device 400 is provided within the imaging range of the imaging device 20. By providing the weight measuring device 400 in this manner, it becomes possible to photograph the target scrap while measuring its weight with the weight measuring device 400. The determination result of the removal target determination device 60 may be transmitted to the terminal device 10. In this case, the person (worker) operating the terminal device 10 may remove the removal target taboo object from the target scrap based on the information indicating the removal target taboo object output on the terminal device 10.

[0125] In the estimation system 100 configured in this manner, the removal target determination device 60 outputs specific examples of the types and amounts of prohibited substances that should be removed so that the amount contained in the target scrap is less than the upper impurity limit. This makes it easy to remove prohibited substances so as to satisfy the requirements for the upper impurity limit.

[0126] [Fourth embodiment] FIG. 27 is a schematic block diagram showing the system configuration of a fourth embodiment of the estimation system 100 of the present invention. The estimation system 100 of the fourth embodiment is the first embodiment of the estimation system 100, and further includes a blending information determination device 70. The blending information determination device 70 determines the blending ratio of scrap from multiple lots whose impurity element concentrations have been estimated as target scrap to satisfy a predetermined standard. By determining the blending ratio with reference to the output of the blending information determination device 70, the person determining the scrap blend can more easily and accurately produce steel using scrap. The fourth embodiment of the estimation system 100 will be described below.

[0127] The blending information determination device 70 determines the blending of scrap from multiple lots. The blending information determination device 70 is configured using an information processing device such as a personal computer or a server device. Figure 28 is a schematic block diagram showing a specific example of the functional configuration of the blending information determination device 70. As shown in the example, the blending information determination device 70 includes a communication unit 71, a memory unit 72, and a control unit 73.

[0128] The communication unit 71 is a communication device. The communication unit 71 may be configured as, for example, a network interface. The communication unit 71 communicates data with other devices via the network 200 in accordance with the control of the control unit 73. The communication unit 71 may be a device that performs wireless communication or a device that performs wired communication.

[0129] The storage unit 72 is configured using a storage device such as a magnetic hard disk drive or a semiconductor storage device. The storage unit 72 stores data used by the control unit 73. The storage unit 72 may function as an inventory information storage unit 721, for example.

[0130] The inventory information storage unit 721 stores inventory information, which is information related to the inventory of iron scrap. FIG. 29 is a diagram showing a specific example of inventory information. The inventory information shown in FIG. 29 has four values: lot identification information, inventory quantity, average unit price, and impurity concentration. The lot identification information is identification information for each lot. The inventory quantity indicates the inventory quantity of iron scrap belonging to each lot. The average unit price indicates the average unit price of iron scrap belonging to each lot. For example, if each lot contains only one target scrap, the amount set for that target scrap is reflected in the average unit price. For example, if each lot contains multiple target scrap, the average unit price is calculated based on the weight and purchase price of the target scrap contained in each lot. The impurity concentration indicates the average concentration of impurity elements contained in the iron scrap belonging to each lot. For example, if each lot contains only one target scrap, the impurity concentration is reflected in the impurity concentration. For example, when each lot contains a plurality of target scraps, the impurity concentration is calculated based on the weight of the target scrap contained in each lot and the estimated values ​​of the impurity element concentrations.

[0131] Next, the control unit 73 will be described. The control unit 73 is configured using one or more hardware processors such as a CPU and one or more memories (main storage devices). The memory is configured using storage devices such as RAM, ROM (Read Only Memory), etc. The control unit 73 functions by one or more hardware processors executing one or more programs stored in the memory to perform various calculations. The control unit 73 functions as, for example, an information control unit 731, a blending information determination unit 732, and an output control unit 733. Note that all or part of the functions of the control unit 73 may be realized using hardware such as an ASIC, PLD, or FPGA. The above programs may be recorded on a computer-readable recording medium. Examples of computer-readable recording media include portable media such as flexible disks, magneto-optical disks, ROMs, CD-ROMs, and semiconductor storage devices (e.g., SSDs), as well as storage devices such as hard disks and semiconductor storage devices built into a computer system. The above programs may be transmitted via telecommunications lines.

[0132] The information control unit 731 acquires information from other devices, such as the terminal device 10 and the estimation device 30. A specific example of the acquired information is information indicating predetermined conditions that must be met when preparing molten steel for a desired steel grade. The predetermined conditions may indicate, for example, the upper limit of impurities (e.g., copper) that can be contained in the molten steel. The predetermined conditions may also include conditions related to price. For example, they may include an upper limit on the price, or a condition regarding the price as "as low as possible." Note that the price in this embodiment is a price that has already been determined based on market conditions, purchasing history, etc., and is different from the price (purchase price) to be used in future transactions between dealers. The information control unit 731 records the acquired information in a storage device such as a memory. The information control unit 731 transmits information to other devices, such as the terminal device 10 and the estimation device 30. A specific example of the transmitted information is the determination result of the blending information. Such information exchange between the information control unit 731 and other devices may be performed, for example, via communication using the communication unit 71.

[0133] The blending information determination unit 732 determines the blending of iron scrap so as to satisfy the specified predetermined conditions. For example, the blending information determination unit 732 may determine the blending using a predetermined algorithm (for example, linear programming) under at least one of the following predetermined conditions (constraint conditions). Specific examples of the predetermined conditions include the following conditions: -Conditions regarding the concentration of impurity elements contained in molten steel (e.g., upper limit) · Requirements regarding the total cost of the total amount of steel scrap contained in the molten steel (e.g., minimization) - Requirements regarding the amount of scrap steel used (e.g., minimization) The more expensive the scrap iron is purchased, the less it should be used. The cheaper the scrap iron is purchased, the more it will be used.

[0134] For example, the blending may be determined as follows using the inventory information shown in FIG. 29, with the target value for the total weight of iron scrap used in the blending and the upper limit value for the impurity (e.g., copper) concentration as constraints. First, the constraints are obtained. As a premise, assume that new scrap is used as a specific example of high-grade scrap (high-grade scrap) to be blended with low-grade scrap (low-grade scrap). Note that high-grade scrap does not need to be limited to new scrap; for example, it may be molten iron obtained from a blast furnace, or other materials may be used. In this case, the average impurity concentration (D high-grade scrap) and average unit price of the high-grade scrap (new scrap) are obtained, and the inventory amount of high-grade scrap is assumed to be unlimited. In this case, the impurity concentration of each lot (e.g., D11, D12, D13) and the usage amount of each lot (e.g., t1, t2, t3) are used to solve the following problem. Note that "C high-grade scrap" indicates the average unit price of high-grade scrap, "D high-grade scrap" indicates the impurity concentration of high-grade scrap, "t high-grade scrap" indicates the weight of high-grade scrap used, and "tn" indicates the weight of each lot n used. Σ(Dn×tn)+D high-quality trash×t high-class trash = (D11 × t1 + D12 × t2 + D13 × t3 + D high-grade scrap × t high-grade scrap) / (weight Σtn + t high-grade scrap) <= upper limit of impurity concentration The constraints are as follows: 1. Weight Σtn+t High-grade scrap meets the target weight 2. Usage price Σ (average unit price n × tn) + C high-grade scrap × t high-grade scrap is the smallest In the above formula, "<=" indicates that the value on the left side is less than or equal to the value on the right side.

[0135] It should be noted that not all of the materials contained in the composition determined by the composition information determination unit 732 necessarily need to be iron scrap. Also, the inventory amount may be obtained based on delivery data (data indicating what was purchased and how much (e.g., tons) was purchased). Also, when a material containing impurities whose exact amount is known is used as a material, the concentration of the impurity elements may be subtracted from the total allowable amount of impurities, and the composition of iron scrap may be determined based on the obtained difference.

[0136] The output control unit 733 outputs information indicating the determination result by the blending information determination unit 732. The output control unit 733 may output the information indicating the determination result by, for example, transmitting it to the terminal device 10. The output control unit 733 may output the information indicating the determination result by controlling an output device connected to the device itself. For example, the output control unit 733 may cause an image display device connected to the device itself to display a character string or an image indicating the determination result. For example, the output control unit 733 may cause a speaker connected to the device itself to output a sound indicating the determination result.

[0137] In the estimation system 100 configured as described above, the blending information determination device 70 can easily determine an appropriate blending for producing molten steel that satisfies predetermined conditions using iron scrap inventory, thereby promoting demand for iron scrap.

[0138] [Fifth embodiment] FIG. 30 is a schematic block diagram showing the system configuration of a fifth embodiment of the estimation system 100 of the present invention. The estimation system 100 of the fifth embodiment further includes a price determination device 80 in addition to the first embodiment of the estimation system 100. The price determination device 80 determines the price of the target scrap based on the estimation result by the estimation device 30. The fifth embodiment of the estimation system 100 will be described below. Note that the price determined by the price determination device 80 is the purchase price (purchase price) of a transaction between one trader and another trader, for example, when one trader brings iron scrap to another trader and requests purchase.

[0139] The price determination device 80 is configured using an information processing device such as a personal computer or a server device. Fig. 31 is a schematic block diagram showing a specific example of the functional configuration of the price determination device 80. As shown in the example, the price determination device 80 includes a communication unit 81, a storage unit 82, and a control unit 83.

[0140] The communication unit 81 is a communication device. The communication unit 81 may be configured as, for example, a network interface. The communication unit 81 communicates data with other devices via the network 200 in accordance with the control of the control unit 83. The communication unit 81 may be a device that performs wireless communication or a device that performs wired communication.

[0141] The storage unit 82 is configured using a storage device such as a magnetic hard disk drive or a semiconductor storage device. The storage unit 82 stores data used by the control unit 83. The storage unit 82 may function as a price information storage unit 821, for example.

[0142] The price information storage unit 821 stores price information, which is information related to the price of iron scrap. FIG. 32 is a diagram showing a specific example of price information. The price information shown in FIG. 32 has two values: impurity concentration and unit price. The impurity concentration indicates the range of the concentration of impurity elements contained per unit weight (e.g., 1 ton) of the target scrap. The unit price indicates the unit price given to target scrap whose impurity concentration falls within the range. By referencing the price information, it is possible to obtain the unit price for each concentration of impurity elements. In other words, by determining the price using the price information, it is possible to determine the price according to the impurity concentration contained in the target scrap.

[0143] Next, the control unit 83 will be described. The control unit 83 is configured using one or more hardware processors such as a CPU and one or more memories (main storage devices). The memory is configured using storage devices such as RAM and ROM (Read Only Memory). The control unit 83 functions by having one or more hardware processors execute one or more programs stored in the memory to perform various calculations. The control unit 83 functions, for example, as an information control unit 831, a price determination unit 832, and an output control unit 833. Note that all or part of the functions of the control unit 83 may be realized using hardware such as an ASIC, PLD, or FPGA. The above programs may be recorded on a computer-readable recording medium. Examples of computer-readable recording media include portable media such as flexible disks, magneto-optical disks, ROMs, CD-ROMs, and semiconductor storage devices (e.g., SSDs), as well as storage devices such as hard disks and semiconductor storage devices built into a computer system. The above programs may be transmitted via telecommunications lines.

[0144] The information control unit 831 acquires information from other devices such as the terminal device 10 and the estimation device 30. A specific example of the acquired information is, for example, the result of estimation of the concentration of impurity elements by the estimation device 30. The information control unit 831 transmits information to other devices such as the terminal device 10 and the estimation device 30. A specific example of the transmitted information is, for example, the result of determining the price of the target scrap. Such exchange of information between the information control unit 831 and other devices may be performed by communication using, for example, the communication unit 81.

[0145] The price determination unit 832 determines the unit price of the target scrap based on the estimation results from the estimation device 30. For example, the price determination unit 832 receives as input the estimation results of the impurity element concentrations from the estimation device 30. The price determination unit 832 obtains the unit price corresponding to the impurity concentration indicated by the input estimation results based on price information such as that shown in FIG. 32. When the price determination unit 832 obtains information regarding the weight of the target scrap from the weight measuring device 400, it may determine the price of the target scrap by multiplying the determined unit price by the weight.

[0146] The output control unit 833 outputs information indicating the determination result by the price determination unit 832. The output control unit 833 may output the information indicating the determination result by, for example, transmitting it to the terminal device 10. The output control unit 833 may output the information indicating the determination result by controlling an output device connected to the device itself. For example, the output control unit 833 may cause an image display device connected to the device itself to display a character string or an image indicating the determination result. For example, the output control unit 833 may cause a speaker connected to the device itself to output a sound indicating the determination result.

[0147] In the estimation system 100 configured in this manner, the price determination device 80 can determine the price according to the concentration of impurities contained in the target scrap.

[0148] FIG. 33 is a diagram illustrating an example of the hardware configuration of an information processing device 90 applied to this embodiment. The information processing device 90 includes a processor 91, a main memory device 92, a communication interface 93, an auxiliary memory device 94, an input / output interface 95, and an internal bus 96. The processor 91, the main memory device 92, the communication interface 93, the auxiliary memory device 94, and the input / output interface 95 are communicably connected to each other via the internal bus 96. The information processing device 90 may be applied to, for example, the terminal device 10, the estimation device 30, the model construction device 40, the taboo substance estimation device 50, the removal target determination device 60, the blending information determination device 70, and the price determination device 80. In this case, for example, the communication unit of each device may be configured using the communication interface 93. For example, the memory unit of each device may be configured using the auxiliary memory device 94. Furthermore, the control unit of each device may be configured using the processor 91 and the main memory device 92.

[0149] Although an embodiment of the present invention has been described above in detail with reference to the drawings, the specific configuration is not limited to this embodiment, and includes designs within the scope of the gist of the present invention. [Explanation of symbols]

[0150] 100... estimation system, 10... terminal device, 20... imaging device, 30... estimation device, 31... communication unit, 32... memory unit, 321... estimation model memory unit, 33... control unit, 331... information control unit, 332... estimation unit, 333... output control unit, 40... model construction device, 41... communication unit, 42... memory unit, 421... known data memory unit, 422... estimation model memory unit, 43... control unit, 431... information control unit, 432... model construction control unit, 50... taboo substance estimation device, 60... removal target determination device, 70... blending information determination device, 80... price determination device

Claims

1. an estimation unit that estimates the concentration of impurity elements contained in the iron scrap to be determined for each lot using an estimation model that is constructed in advance using known data including supplementary information that is information about the iron scrap containing iron to be recycled and information about the concentration of impurity elements contained in the iron scrap; and a blending information determination unit that determines the blending of the lot of iron scrap based on the concentration of impurity elements contained in the iron scrap for each lot estimated by the estimation unit and the price of the iron scrap for each lot, so that the iron scrap used for blending satisfies at least one of predetermined constraints including a condition regarding the concentration of impurity elements contained in the iron scrap, a condition regarding the total cost of the total amount of iron scrap, and a condition regarding the amount of iron scrap used.

2. An estimation system having an estimation unit and a blending information determination unit is used, an estimation step in which the estimation unit estimates the concentration of impurity elements contained in the iron scrap to be determined for each lot using an estimation model previously constructed using known data including supplementary information that is information about the iron scrap containing iron to be recycled and information about the concentration of impurity elements contained in the iron scrap; and a blending information determination step of determining the blending of the lot of iron scrap based on the concentration of impurity elements contained in the iron scrap for each lot estimated in the estimation step and predetermined conditions related to the iron scrap for each lot, so that the iron scrap used for blending satisfies predetermined constraints consisting of at least one of conditions related to the concentration of impurity elements contained in the iron scrap, conditions related to the total cost of the total amount of iron scrap, and conditions related to the amount of iron scrap used.

3. an estimation unit that estimates the concentration of impurity elements contained in the iron scrap to be determined for each lot using an estimation model that is constructed in advance using known data including supplementary information that is information about the iron scrap containing iron to be recycled and information about the concentration of impurity elements contained in the iron scrap; and a blending information determination unit that determines the blending of the lot of iron scrap based on the concentration of impurity elements contained in the iron scrap for each lot estimated by the estimation unit and the price of the iron scrap for each lot, so that the iron scrap used for blending satisfies at least one of predetermined conditions including a condition regarding the concentration of impurity elements contained in the iron scrap, a condition regarding the total cost of the total amount of iron scrap, and a condition regarding the amount of iron scrap used.

Images