Setting support system, setting support method, setting support device, and program

Abstract

A setting assistance system according to one aspect of the present disclosure comprises: an acquisition unit which acquires mixture information pertaining to a mixture, which includes a plurality of types of objects, from a first sorting device that sorts the mixture for each type of the objects, and operation condition information about the first sorting device; a reception unit which receives mixture information about the mixture that is sorted by a second sorting device; a similarity calculation unit which calculates similarity between the mixture information received by the reception unit and the mixture information acquired by the acquisition unit; and a setting unit which sets an operation condition to the second sorting device on the basis of the similarity calculated by the similarity calculation unit and the operation condition information about the first sorting device.

Description

[Technical field]

[0001] The present disclosure relates to a setting assistance system, a setting assistance method, a setting assistance device, and a program. [Background technology]

[0002] Conventionally, there is known a technique for separating a mixture containing a plurality of types of objects into different types of objects. In this separating technique, for example, there is an electrostatic separating device that electrically charges a plurality of types of objects contained in the mixture and separates specific objects as objects to be recycled by electrostatic separation. As a technique for improving the accuracy of electrostatic separation, for example, an electrostatic separating device described in the following Patent Document 1 is known.

[0003] The electrostatic sorting device of Patent Document 1 stirs a group of plastic pieces containing a mixture of multiple types of plastic pieces, charges each plastic piece, drops each charged plastic piece into an electric field, and collects the dropped plastic pieces in a collection box. A movable partition plate is provided in the collection box. This electrostatic sorting device samples a portion of the group of plastic pieces to measure the composition ratio of the multiple types of plastic pieces, analyzes the trend of change in the composition ratio from the composition ratio measured in the past and the composition ratio measured this time, and controls the position of the partition plate based on the analysis results. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] International Publication No. 2021-166178 Summary of the Invention [Problem to be solved by the invention]

[0005] In sorting objects, there is a trade-off between the recovery rate of a particular object and the purity of the particular object, and it is required to maximize the recovery rate while satisfying the purity desired by the user, for example. The recovery rate and purity recovery results vary depending on the state of the mixture fed into the sorting device. The state of the mixture is indicated by, for example, the type of object to be sorted, the amount fed, the mixture ratio (composition ratio), and the particle size. The state of the mixture differs depending on the product to be recycled, pre-processing, season, input lot, etc.

[0006] For example, when a sorting device is newly installed, it is not possible to set appropriate sorting conditions for the device because the relationship between the state of the mixture and the recovery results is unknown. Even after the sorting device is installed, it is difficult to stabilize the recovery results because the recovery results may change due to fluctuations in the state of the mixture.

[0007] The present disclosure has been made in consideration of the above circumstances, and aims to provide a setting support system, a setting support method, a setting support device, and a program that can assist in appropriately setting operating conditions for a sorting device. [Means for solving the problem]

[0008] The present disclosure has been made to solve the above-mentioned problems, and one aspect of the present disclosure is a setting support system including: an acquisition unit that acquires mixture information regarding a mixture in a first sorting device that sorts a mixture containing multiple types of objects by type of the objects and operating condition information of the first sorting device; a reception unit that accepts mixture information of a mixture to be sorted by a second sorting device; a similarity calculation unit that calculates a similarity between the mixture information accepted by the reception unit and the mixture information acquired by the acquisition unit; and a setting unit that sets operating conditions for the second sorting device based on the similarity calculated by the similarity calculation unit and the operating condition information of the first sorting device.

[0009] Another aspect of the present disclosure is a setting assistance method including the steps of: acquiring mixture information on a mixture in a first sorting device that sorts a mixture containing multiple types of objects by type of the objects and operating condition information of the first sorting device; accepting mixture information on a mixture to be sorted by a second sorting device; calculating a similarity between the mixture information on the mixture to be sorted by the second sorting device and the mixture information on the mixture in the first sorting device; and setting operating conditions for the second sorting device based on the similarity and the operating condition information of the first sorting device.

[0010] Another aspect of the present disclosure is a setting assistance device comprising: a similarity calculation unit that calculates a similarity between mixture information regarding a mixture in a first sorting device that sorts a mixture containing multiple types of objects by type of the objects, operating condition information of the first sorting device, and mixture information of a mixture to be sorted by a second sorting device; and a setting unit that sets operating conditions for the second sorting device based on the similarity calculated by the similarity calculation unit and the operating condition information of the first sorting device.

[0011] Another aspect of the present disclosure is a program that causes a computer of a setting assistance device to execute the steps of calculating a similarity between mixture information regarding a mixture in a first sorting device that sorts a mixture containing multiple types of objects by type of the objects, operating condition information of the first sorting device, and mixture information of a mixture to be sorted by a second sorting device, and setting operating conditions for the second sorting device based on the similarity and the operating condition information of the first sorting device. Effect of the Invention

[0012] According to the present disclosure, it is possible to assist in setting appropriate operating conditions for a sorting device. [Brief description of the drawings]

[0013] [Figure 1] 1 is a block diagram showing an example of a setting support system according to a first embodiment. [Diagram 2] 4 is a flowchart showing an example of an operation procedure of the setting support system in the first embodiment. [Diagram 3] 4 is a sequence diagram showing an example of processing of the setting support system in the first embodiment. FIG. [Figure 4] FIG. 1 is a diagram illustrating an example of an electrostatic separation device according to a first embodiment. [Diagram 5] FIG. 2 is a block diagram showing an example of a functional configuration of the electrostatic separation device according to the first embodiment. [Figure 6] FIG. 4 is a diagram illustrating an example of an operation of a similarity calculation unit in the first embodiment. [Figure 7] FIG. 4 is a diagram illustrating an example of an operation of a similarity calculation unit in the first embodiment. [Figure 8] FIG. 11 is an explanatory diagram showing an example of an operation of selecting a trained model by the setting support system in the first embodiment and controlling the electrostatic separation device. [Figure 9] FIG. 11 is an explanatory diagram showing an example of an operation of selecting a trained model by the setting support system in the first embodiment and controlling the electrostatic separation device. [Figure 10] FIG. 11 is a block diagram showing an example of a setting support system according to a second embodiment. [Figure 11] FIG. 11 is a sequence diagram showing an example of processing of the setting support system in the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] A setting support system, a setting support method, a setting support device, and a program to which the present invention is applied will be described below with reference to the drawings.

[0015] FIG. 1 is a block diagram showing an example of a setting support system 1 in the first embodiment. The setting support system 1 includes a setting support device 100, a plurality of electrostatic separation devices 200A and 200B, a database device 300, a learning device 400, and a control device 410. The setting support device 100, the electrostatic separation devices 200A and 200B, the learning device 400, and the control device 410 each have a communication interface (not shown) such as a NIC (Network Interface Card) or a wireless communication module for connecting to a network such as the Internet. The network may include, for example, a general-purpose network such as the Internet, and a private network such as local 5G or Wifi (registered trademark).

[0016] In the embodiment, an electrostatic sorting device that uses static electricity to sort recyclable materials from a mixture containing multiple types of resins is described as a sorting device that sorts a mixture containing multiple types of objects by object type, but is not limited to this. The setting support system 1 of the embodiment can also be applied to a sorting device that sorts a specific object from a mixture containing metal, and can also be applied to a sorting device that sorts a specific object from a mixture containing food.

[0017] The electrostatic separation device 200A is, for example, a new electrostatic separation device that does not perform electrostatic separation, and the electrostatic separation device 200B is, for example, an existing electrostatic separation device that performs electrostatic separation. The setting support system 1 shows one electrostatic separation device 200A and multiple electrostatic separation devices 200B as shown in FIG. 1, and describes setting the operating conditions of the new electrostatic separation device 200A based on the operating condition information of the existing electrostatic separation device 200B, but is not limited to this. The operating conditions of the other electrostatic separation device 200B may be set based on the operating condition information of the existing electrostatic separation device 200B. In the following description, when the electrostatic separation devices 200A and 200B are collectively referred to, they will simply be described as "electrostatic separation device 200".

[0018] The electrostatic sorting device 200 operates to sort materials to be recycled from a mixture containing multiple types of resins. The materials to be recycled include multiple types of pieces to be sorted that have different charging characteristics. The electrostatic sorting device 200 includes a detection unit 202 that detects mixture information of the mixture input into the electrostatic sorting device 200. The electrostatic sorting device 200 transmits the mixture information and the operating condition information to the setting support device 100. Note that the electrostatic sorting device 200 can also be applied to a recycling facility that sorts materials to be sorted from raw materials by other methods such as a specific gravity method instead of an electrostatic method.

[0019] The mixture information is information about the mixture in the electrostatic separation device 200. For example, the mixture information is information including at least one of the type, composition, and particle size of the objects.

[0020] The operating condition information is information indicating the operating conditions of the electrostatic separation device 200. The operating condition information may be the magnitude of the voltage applied to the electrode pair that applies an electrostatic force to a charged object, or the position of a partition that divides an area in which an object induced by an electrostatic force is accommodated. The operating condition information may be information based on a trained model that is trained using mixture information, operating conditions, and collection results as learning data and is trained to output operating conditions that result in optimal collection results when mixture information is input. A trained model is created for each mixture information. The operating condition information may be information based on one trained model that combines multiple trained models based on similarity. The information based on the trained model may be, for example, information for setting the trained model, or may be information output from the trained model. The information for setting the trained model is information for allowing the control device 410 of another electrostatic separation device 200 to use the trained model of a certain electrostatic separation device 200.

[0021] The learning device 400 acquires mixture information, operating condition information, and recovery results from the electrostatic separation device 200B as learning data. The recovery results are, for example, the recovery rate of a specific object and the purity of the specific object. The recovery results may include the environmental load of the electrostatic separation device 200, the operating cost of the electrostatic separation device 200, and the operating speed. The learning device 400 constructs a trained model 402 that outputs operating conditions that result in optimal recovery results when the mixture information is input. The learning device 400 constructs the trained model 402 by updating processing parameters such as a filter (also called a weight or bias) included in a neural network. The machine learning method in the trained model 402 may be commonly used deep learning or random forest.

[0022] The control device 410 is an information processing device that controls the electrostatic separation device 200 using the trained model 402. The control device 410 inputs the mixture information detected by the detection unit 202 to the trained model 402, and controls the electrostatic separation device 200 based on the operating condition information output from the trained model 402.

[0023] The database device 300 is a device including a data storage device, a processor for constructing a database (DB), and the like. The database device 300 stores, for example, a user database 310 and information on the trained model 320. The user database 310 is, for example, a database in which user information for identifying a user or the electrostatic separation device 200, mixture information, and the trained model 320 are associated with each other. The trained model 320 corresponds to the trained model 402 constructed by the training device 400. The trained model 320 is stored in the database device 300 for each mixture. The trained model 320 may be stored in the database device 300 by type, composition, or particle size of the resin pieces.

[0024] The database device 300 acquires mixture information and information for setting a trained model from the electrostatic separation device 200, the learning device 400, or the control device 410 via a network. The database device 300 acquires user information based on the network or a user's operation. In this way, the database device 300 creates and updates a user database 310. In addition, the database device 300 acquires information for setting a trained model 402 from the learning device 400, and stores a trained model 320 corresponding to the mixture information. The database device 300 may include a learning device, acquire mixture information, operating condition information, and collection results of the electrostatic separation device 200, and use the acquired information to construct a trained model.

[0025] The database device 300 may store the mixture information and the operating condition information in the electrostatic separation device 200B in association with each other. The database device 300 transmits the mixture information and the operating condition information in the electrostatic separation device 200B to the setting support device 100 in response to a request from the setting support device 100.

[0026] The setting support device 100 is an information processing device in which a processor such as a CPU (Central Processing Unit) executes a program stored in a program memory. The setting support device 100 includes, for example, an acquisition unit 102, a reception unit 104, a similarity calculation unit 106, and a setting unit 108. The acquisition unit 102, the reception unit 104, the similarity calculation unit 106, and the setting unit 108 are functional units realized by the processor executing the program.

[0027] The acquiring unit 102 acquires mixture information related to the mixture in the electrostatic separation device 200B that separates a mixture containing a plurality of types of objects into individual object types, and operating condition information of the electrostatic separation device 200B. The accepting unit 104 accepts the mixture information of the mixture to be separated by the electrostatic separation device 200A.

[0028] The similarity calculation unit 106 calculates the similarity between the mixture information accepted by the acceptance unit 104 and the mixture information acquired by the acquisition unit 102. The setting unit 108 sets operating conditions for the electrostatic separation device 200A based on the similarity calculated by the similarity calculation unit 106 and the operating condition information of the electrostatic separation device 200B.

[0029] The setting unit 108 may, for example, set the operating condition information of the electrostatic separation device 200B in the electrostatic separation device 200A by transmitting the operating condition information of the electrostatic separation device 200B to the electrostatic separation device 200A. The setting unit 108 may, for example, set the operating condition information of the electrostatic separation device 200B in the electrostatic separation device 200A by transmitting the operating condition information of the electrostatic separation device 200B to a terminal of a person who sets up the electrostatic separation device 200A. The setting unit 108 may, for example, display the operating condition information of the electrostatic separation device 200B on a terminal of a person who sets up the electrostatic separation device 200A.

[0030] In the first embodiment, the acquisition unit 102, the reception unit 104, the similarity calculation unit 106, the setting unit 108, the detection unit 202, the learning device 400, and the control device 410 are distributed as shown in FIG. 1, but are not limited to this and may be distributed or consolidated in other forms. Each unit, such as the acquisition unit 102, the reception unit 104, the similarity calculation unit 106, and the setting unit 108, may be distributed to multiple information processing devices. For example, the similarity calculation unit 106 and the setting unit 108 may be provided in one setting support device 100, and the acquisition unit 102 and the reception unit 104 may be provided in a device other than the setting support device 100.

[0031] FIG. 2 is a flowchart showing an example of an operation procedure of the setting support system 1 in the first embodiment. First, the acquisition unit 102 acquires mixture information of the existing electrostatic separation device 200B (step S100). Next, the acquisition unit 102 acquires operating condition information of the existing electrostatic separation device 200B (step S102). The acquisition unit 102 acquires the mixture information of the electrostatic separation device 200B and stores it in the user database 310. The acquisition unit 102 acquires a trained model 402 as operating condition information, and stores it as a trained model 320 in association with the mixture information of the electrostatic separation device 200B. Next, the reception unit 104 accepts the mixture information of the new electrostatic separation device 200A (step S104). Next, the similarity calculation unit 106 calculates the similarity between the mixture information of the electrostatic separation device 200A accepted by the reception unit 104 and the mixture information of the electrostatic separation device 200B acquired by the acquisition unit 102 (step S106).

[0032] Next, the setting unit 108 selects an optimal trained model 320 from the trained models 320 based on the similarity calculated by the similarity calculation unit 106 (step S108). The setting unit 108 determines the operating conditions of the electrostatic separation device 200A using the selected trained model 320 (step S110). The setting unit 108 sets the determined operating conditions to the new electrostatic separation device 200A (step S112).

[0033] FIG. 3 is a sequence diagram showing an example of processing of the setting support system 1 in the first embodiment. The existing electrostatic separation device 200B transmits mixture information and information S1 for setting the trained model 402 to the setting support device 100. The acquisition unit 102 receives the information S1 related to the mixture information and the trained model 402 from the electrostatic separation device 200B, and transmits an information registration request S2 to the database device 300. As a result, the database device 300 updates the user database 310 based on the mixture information and the information S1 related to the trained model 402, and stores the trained model 320. The information related to the trained model 402 is information for setting the trained model 402, but is not limited thereto, and may be an operating condition output from the trained model 402.

[0034] When setting operating conditions for a new electrostatic separation device 200A, the electrostatic separation device 200A transmits mixture information S3 to the setting support device 100, and the reception unit 104 receives the mixture information S3. The similarity calculation unit 106 calculates the similarity and transmits a request S4 for information on the trained model 320 selected based on the similarity. The database device 300 transmits information S5 on the selected trained model to the setting support device 100 in accordance with the request S4. The setting unit 108 transmits operating condition information S6 to the electrostatic separation device 200A based on information S5 on the trained model transmitted from the database device 300. The setting unit 108 may create operating condition information for the new electrostatic separation device 200A using the trained model and transmit the information to the electrostatic separation device 200A. The setting unit 108 may transmit information for setting the selected trained model to the electrostatic separation device 200A.

[0035] FIG. 4 is a diagram showing an example of an electrostatic sorting device 200 in the first embodiment. The electrostatic sorting device 200 includes, for example, a supply unit 220 that supplies a mixture 210 containing multiple types of plastic pieces A and B to be sorted, a charging unit 230 that charges the plastic pieces A and B, an electric field generating unit 240 that generates an electrostatic field, and a collection box 250 that collects the plastic pieces A and B. A control device 410 that controls each unit in the electrostatic sorting device 200 and performs various calculations is connected to the electrostatic sorting device 200. The mixture 210 in the embodiment includes plastic pieces containing a mixture of multiple types of materials obtained by crushing the housings of home appliances such as air conditioners, refrigerators, and washing machines to be recycled. The plastic pieces A and B in the embodiment are ABS (Acrylonitrile Butadiene Styrene) pieces and PS (Polystyrene) pieces having different charging characteristics. The plastic pieces A and B are, for example, about 5 mm square in size. In FIG. 4, the ABS piece as the plastic piece A is shown in white, and the PS piece as the plastic piece B is shown in solid.

[0036] The supply unit 220 includes, for example, a hopper 221 and an input feeder 222. The hopper 221 stores plastic pieces A and B that have been dried by a dryer (not shown). The hopper 221 supplies a predetermined amount of plastic pieces A and B per unit time to the input feeder 222. The input feeder 222 inputs the plastic pieces A and B supplied from the hopper 221 into the charging unit 230.

[0037] The charging unit 230 includes, for example, a charging cylinder 231 and a vibration feeder 232. The charging cylinder 231 agitates the plastic pieces A and B by rotating. The plastic pieces A and B rub ​​against each other inside the charging cylinder 231 and become charged due to rotational agitation. The charged plastic pieces A and B each have a polarity and a charge amount according to the triboelectric series. Specifically, the plastic piece A is positively charged, and the plastic piece B is negatively charged.

[0038] The electrically charged plastic pieces A and B are discharged into the vibrating feeder 232. The positively charged plastic piece A and the negatively charged plastic piece B are attracted to each other by electrostatic force. The vibrating feeder 232 pushes the plastic pieces A and B forward (+X side) while vibrating them up and down (Z direction), causing the plastic pieces A and B to fall from the tip of the vibrating feeder 232 while being separated from each other.

[0039] The electric field generating unit 240 includes, for example, a pair of electrodes 241 and 242 and a DC power supply 243. Each of the pair of electrodes 241 and 242 is formed, for example, in a flat plate shape. The pair of electrodes 241 and 242 are arranged opposite each other in the X-axis direction, sandwiching the path along which the plastic pieces A and B fall. A ground voltage GND is applied to the electrode 241. The DC power supply 243 generates an electrostatic field between the pair of electrodes 241 and 242 by applying a predetermined DC voltage between the electrodes 241 and 242. The plastic pieces A and B that have fallen from the vibration feeder 232 fall while being attracted to one of the pair of electrodes 241 and 242 by an electrostatic force according to the charge state such as polarity and charge amount. In other words, the plastic pieces A and B fall to different positions according to the charge state. In the embodiment, the plastic piece A is positively charged and is attracted to the electrode 241 side and falls. The plastic piece B is negatively charged and is attracted to the electrode 242 side and falls.

[0040] The dropped plastic pieces A and B are collected in collection box 250. Collection box 250 includes, for example, housing part 251 and partition plates 252 and 253. In collection box 250, area 250A surrounded by housing part 251 and partition plate 252, area 250B surrounded by housing part 251 and partition plate 253, and area 250C sandwiched between partition plates 252 and 253 are formed.

[0041] The partition plates 252 and 253 are movable in, for example, the X direction in the figure. The partition plates 252 and 253 are connected to, for example, the driving units 254 and 255. The driving units 254 and 255 are, for example, motors that generate driving forces that move the partition plates 252 and 253. The control device 410 supplies control signals to the driving units 254 and 255. The partition plates 252 and 253 move to adjust the sizes of the areas 250A, 250B, and 250C. The partition plates 252 and 253 may be moved manually based on the positions of the partition plates 252 and 253 calculated by the control device 410. In the embodiment, the partition plates 252 and 253 are described as partitions that divide the areas 250A, 250B, and 250C that contain objects induced by electrostatic force, but are not limited to a plate shape and may have other shapes.

[0042] Plastic pieces A and B dropped from vibrating feeder 232 are collected in either area 250A, 250B, or 250C depending on their charge state. Positively charged plastic pieces A are collected mainly in area 250A. Negatively charged plastic pieces B are collected mainly in area 250B. Plastic pieces that are not sufficiently charged are collected in area 250C. The plastic pieces collected in area 250A, area 250B, and area 250C are transported by a conveyor (not shown) and stored in separate containers.

[0043] The electrostatic separation device 200 may return the plastic pieces collected in the area 250C to the supply section 220 and perform separation again.

[0044] 5 is a block diagram showing an example of a functional configuration of the electrostatic sorting device 200 in the first embodiment. The electrostatic sorting device 200 includes, for example, a data processing unit 261, an input amount conversion unit 262, a data processing unit 263, image processing units 264, 265, 266, and 267, an input plastic sensor unit 270, and a recovered plastic sensor unit 280. The input plastic sensor unit 270, the input amount conversion unit 262, the data processing unit 263, and the image processing units 264 and 265 correspond to the detection unit 202 described above.

[0045] The control device 410 outputs operating condition information, such as partition plate position information, voltage information, MID circulation information, charged cylinder information, and feeder speed information, to the data processing unit 261. The partition plate position information is information indicating the positions of the partition plates 252 and 253. The voltage information is information indicating the voltage between the electrodes 241 and 242, and the MID circulation information is information indicating the amount of plastic pieces circulated from the region 250C to the supply unit 220. The charged cylinder information is information indicating the rotation speed of the charged cylinder 231. The feeder speed information is information indicating the vibration speed of the vibration feeder 232. The data processing unit 261 performs, for example, a predetermined data processing, and transmits to the learning device 400 operating condition data in which the partition plate position information, voltage information, MID circulation information, charged cylinder information, and feeder speed information are linked for each time.

[0046] The input plastic sensor unit 270 includes, for example, a particle sensor, a charge sensor, a near-infrared camera, and a visible light camera. The input plastic sensor unit 270 outputs particle passing information detected by the particle sensor for the mixture 210 input to the charging unit 230 to the input amount conversion unit 262. The input amount conversion unit 262 transmits input amount data indicating the input amount of the mixture 210 based on the particle passing information to the learning device 400. The input plastic sensor unit 270 outputs particle charge amount information detected by the charge amount sensor to the data processing unit 263. The data processing unit 263 transmits charge amount data indicating the charge amount of the mixture 210 based on the charge amount information to the learning device 400. The input plastic sensor unit 270 outputs a near-infrared image captured by the near-infrared camera to the image processing unit 264. The image processing unit 264 acquires the composition ratio based on the near-infrared image, and transmits the composition ratio data of the mixture 210 to the learning device 400. The input plastic sensor unit 270 outputs a visible image captured by the visible light camera to the image processing unit 265. The image processing unit 265 obtains information indicating the particle size and foreign matter contained in the mixture 210 based on the visible image, and transmits the particle size and foreign matter data of the mixture 210 to the learning device 400.

[0047] The collected plastic sensor unit 280 includes, for example, a visible light camera, an optical sensor, and a near-infrared camera. The collected plastic sensor unit 280 outputs a visible image of the plastic pieces collected in the collection box 250, detected by the visible light camera or the optical sensor, to the image processing unit 266. The image processing unit 266 acquires the collection rate for each of the plastic pieces A and B based on the visible image, and transmits the collection rate data to the learning device 400. The collected plastic sensor unit 280 outputs a near-infrared image of the plastic pieces collected in the collection box 250, detected by the near-infrared camera, to the image processing unit 267. The image processing unit 267 acquires the purity based on the near-infrared image, and transmits the purity data of the mixture 210 to the learning device 400.

[0048] The learning device 400 uses mixture information including input amount data, charge amount data, composition ratio data, and particle size / foreign matter data, operating condition data, and recovery results including recovery rate data and purity data to train the trained model 402. This enables the learning device 400 to update the trained model.

[0049] When the detection unit 202 detects that another mixture 210 has been introduced into the electrostatic separation device 200 based on the mixture information, the learning device 400 may train a trained model corresponding to the different mixture 210. This allows the learning device 400 to construct a trained model for each mixture 210. The control device 410 can switch the trained model based on the mixture information detected by the detection unit 202.

[0050] When the mixture information detected by the detection unit 202 differs from the mixture information corresponding to the trained model, the control device 410 may transmit the mixture information detected by the detection unit 202 to the setting support device 100. The similarity calculation unit 106 calculates the similarity based on the mixture information acquired from the control device 410, and the setting unit 108 selects the trained model with the highest similarity. The setting unit 108A can transmit information on the selected trained model to the control device 410. The control device 410 can control the electrostatic separation device 200 by switching to a trained model corresponding to the mixture information. This allows the setting unit 108 to change the operating conditions set in the electrostatic separation device 200A based on the similarity.

[0051] FIG. 6 is a diagram showing an example of the operation of the similarity calculation unit 106 in the first embodiment. The similarity calculation unit 106 calculates the similarity between the mixture information of the new electrostatic separation device 200A and the mixture information of the existing electrostatic separation device 200B. The similarity is calculated using at least one or more pieces of information from among the mixing ratio, particle size, and input amount among parameters that affect the type of plastic pieces and the separation efficiency. For example, as shown in the user database 310 in FIG. 6, the similarity calculation unit 106 extracts devices B, E, and G in which the type of plastic pieces contained in the mixture 210 in the new electrostatic separation device 200A matches the type of plastic pieces contained in the mixture 210 in the existing electrostatic separation device 200B (devices A to G).

[0052] 7 is a diagram showing an example of the operation of the similarity calculation unit 106 in the first embodiment. The similarity calculation unit 106 calculates the similarity between the average composition ratio, average particle size, and average input amount in the mixture information of the new electrostatic separation device 200A and the average composition ratio, average particle size, and average input amount in the mixture information of each of the devices B, E, and G. As a result, the similarity calculation unit 106 selects the device G that is most similar to the mixture information of the electrostatic separation device 200A from the mixture information of the devices B, E, and G. At this time, the similarity calculation unit 106 may calculate the similarity by calculating the Euclidean distance or the cosine similarity.

[0053] FIG. 8 is an explanatory diagram showing an example of an operation of selecting a trained model and controlling the electrostatic separation device 200 by the setting support system 1 in the first embodiment. Assume that the database device 300 stores mixture information and trained model A of device A, mixture information and trained model B of device B, and mixture information and trained model C of device C. The similarity calculation unit 106 calculates the similarity for each of device A, device B, and device C by comparing the composition, average composition ratio, composition fluctuation, and average particle size of the mixture information of the new electrostatic separation device 200A with the composition, average composition ratio, composition fluctuation, and average particle size of the mixture information stored in the database device 300. The similarity calculation unit 106 selects the trained model B of device B because the similarity of device B is the highest among devices A, B, and C.

[0054] The setting unit 108 acquires information about the selected trained model B from the database device 300, and inputs the mixture information of the new electrostatic separation device 200A to the trained model B. The setting unit 108 transmits the operating condition information, such as the partition plate position information, voltage information, charged cylinder information, and feeder speed information, output from the trained model B, to the new electrostatic separation device 200A. This allows the electrostatic separation device 200A to perform electrostatic separation according to the partition plate position information, voltage information, charged cylinder information, and feeder speed information transmitted from the setting unit 108.

[0055] FIG. 9 is an explanatory diagram showing an example of an operation of selecting a trained model by the setting support system 1 in the first embodiment and controlling the electrostatic separation device 200. The setting unit 108A may select a plurality of trained models based on the similarity, and the operating condition information may be information based on one trained model obtained by combining the selected plurality of trained models. The setting unit 108 creates one trained model D by combining two trained models B and C that have high similarity among the plurality of similarities calculated by the similarity calculation unit 106. The setting unit 108 inputs mixture information of the new electrostatic separation device 200A to the trained model D, and transmits the operating condition information output from the trained model D to the new electrostatic separation device 200A. The setting unit 108 may transmit information for combining the trained model B and the trained model C to the electrostatic separation device 200A and setting the one trained model D as the trained model 402 in the electrostatic separation device 200A.

[0056] When there is mixture information whose similarity to the mixture information of the electrostatic separation device 200A is equal to or greater than a threshold, the setting unit 108 may transmit information based on a trained model corresponding to the mixture information with the highest similarity to the electrostatic separation device 200A. When there is no mixture information whose similarity to the mixture information of the electrostatic separation device 200A is equal to or greater than a threshold, the setting unit 108 may create a trained model by combining multiple trained models. For example, when the average composition ratio in the mixture information of the electrostatic separation device 200A is 30-70%, a trained model corresponds to the mixture information whose average composition ratio is 30-50%, and another trained model corresponds to the mixture information whose average composition ratio is 50-70%, a trained model corresponding to the mixture information whose average composition ratio is 30-70% can be constructed by combining the two trained models. This allows the setting unit 108 to increase the similarity between the mixture information corresponding to the combined trained model and the mixture information of the electrostatic separation device 200A. As a result, the setting unit 108A can set the operating condition information output from the trained model with a high degree of similarity to the electrostatic separation device 200A, and can cause the trained model with a high degree of similarity to be set to the electrostatic separation device 200A.

[0057] As described above, according to the setting support system 1 of the first embodiment, the mixture information and operating condition information of the electrostatic separation device 200B are acquired, the mixture information of the mixture 210 to be sorted by the electrostatic separation device 200A is received, and the operating conditions of the electrostatic separation device 200A can be set based on the similarity between the mixture information of the electrostatic separation device 200A and the mixture information of the electrostatic separation device 200B and the operating condition information of the electrostatic separation device 200B. As a result, according to the setting support system 1, it is possible to support the setting of appropriate operating conditions for the electrostatic separation device 200A. As a result, according to the setting support system 1, the operating conditions can be determined in a short period of time without the user acquiring data indicating the relationship between the mixture information and the collection result for a new electrostatic separation device 200A.

[0058] Furthermore, the setting support system 1 can calculate the similarity between the mixture information detected by the detection unit 202 and the mixture information of the electrostatic separation device 200B, and change the operating conditions set in the electrostatic separation device 200A based on the similarity. As a result, the setting support system 1 can stabilize the recovery results even if the state of the mixture 210 changes after the electrostatic separation device 200 is introduced.

[0059] The second embodiment will be described below. The second embodiment differs from the first embodiment in that the operating conditions are set for the electrostatic separation device 200A based on the requirements of the customer who owns the electrostatic separation device 200A, the similarity between the mixture information of the electrostatic separation device 200A and the mixture information of the electrostatic separation device 200B, and the operating condition information of the electrostatic separation device 200B. Note that the same parts as those in the first embodiment are denoted by the same reference numerals and will not be described.

[0060] FIG. 10 is a block diagram showing an example of the setting support system 1A in the second embodiment. The setting support system 1A in the second embodiment includes a user terminal device 500. The user terminal device 500 is an information processing device such as a personal computer that performs various processes based on the operation of a user as a customer. The user terminal device 500 transmits a request regarding the electrostatic separation device 200A to the setting support device 100A in response to the operation of the user. The request regarding the electrostatic separation device 200A is, for example, a request to improve the recovery rate of a specific type of object in the mixture 210, a request to improve the purity of a specific type of object in the mixture 210, a request to reduce the environmental load of the electrostatic separation device 200A, a request to reduce the operating cost of the electrostatic separation device 200A, and the like. The environmental load is, for example, a value representing the load on the environment caused by the operation of the electrostatic separation device 200A, such as CO2 emissions. The operating cost is, for example, a cost incurred by the operation of the electrostatic separation device 200A, such as the power consumption of the electrostatic separation device 200A.

[0061] In the setting support device 100A, the reception unit 104 receives a request from a customer who uses the electrostatic separation device 200A from the user terminal device 500. The setting unit 108A sets the operating conditions of the electrostatic separation device 200A based on the request received by the reception unit 104, the mixture information of the electrostatic separation device 200A, the similarity with the mixture information of the electrostatic separation device 200B acquired by the acquisition unit 102, and the operating condition information of the electrostatic separation device 200B. In this way, the setting unit 108A sets the operating conditions that prioritize at least one of the recovery rate, purity, environmental load, and operating cost in response to the request.

[0062] For example, when the receiving unit 104 receives a request to prioritize purity, the setting unit 108A may select the trained model with the highest purity from among trained models whose similarity is equal to or higher than a reference level. The setting unit 108A can adjust the operating conditions in response to the request by inputting mixture information of the electrostatic separation device 200A to the selected trained model and setting the operating conditions based on the output of the trained model to the electrostatic separation device 200A.

[0063] The setting support device 100A includes a billing processing unit 110 that generates billing information based on the setting of operating conditions that satisfy the requirements by the setting unit 108A. The billing processing unit 110 may generate the billing information based on a change in at least one of the recovery rate, purity, environmental load, and operating cost due to the electrostatic separation device 200A operating using the operating conditions that satisfy the requirements set by the setting unit 108A. The billing processing unit 110 may generate the billing information when the operating conditions that have been changed in response to the acceptance of a change in the mixture information or a change in the request by the acceptance unit 104 are provided.

[0064] FIG. 11 is a sequence diagram showing an example of processing of the setting support system 1A in the second embodiment. The existing electrostatic separation device 200B transmits information S1 relating to the mixture information and the trained model 402 to the setting support device 100A. The acquisition unit 102 receives the information S1 relating to the mixture information and the trained model 402 from the electrostatic separation device 200B, and transmits an information registration request S2 to the database device 300. As a result, the database device 300 updates the user database 310 based on the mixture information and the information S1 relating to the trained model 402, and stores the trained model 320.

[0065] When setting operating conditions for a new electrostatic separation device 200A, the electrostatic separation device 200A transmits mixture information S3 to the setting support device 100A, and the reception unit 104 receives the mixture information S3. The similarity calculation unit 106 calculates the similarity and transmits a request S4 for information on the trained model 320 selected based on the similarity. The database device 300 transmits information S5 on the selected trained model to the setting support device 100A in accordance with the request S4. The setting unit 108A transmits operating condition information S6 to the electrostatic separation device 200A based on information S5 on the trained model transmitted from the database device 300. The setting unit 108A may create operating condition information for the new electrostatic separation device 200A using the trained model and transmit it to the electrostatic separation device 200A. The setting unit 108A may transmit information for setting the selected trained model to the electrostatic separation device 200A.

[0066] The user terminal device 500 transmits a user request S7 regarding the electrostatic separation device 200A to the setting support device 100A based on a user operation. The reception unit 104 receives the user request S7. The user request S7 may be a change in mixture information or a change in request. The change in request is, for example, changing the target values ​​of the recovery rate of a specific type of object in the mixture, the purity of a specific type of object in the mixture, the environmental load of the electrostatic separation device 200A, and the operating cost of the electrostatic separation device 200A, or changing a prioritized operating condition among the recovery rate of a specific type of object in the mixture, the purity of a specific type of object in the mixture, the environmental load of the electrostatic separation device 200A, and the operating cost of the electrostatic separation device 200A. The setting unit 108A transmits a model request S8 for acquiring a trained model according to the user request S7 to the database device 300, and acquires information S9 regarding the trained model from the database device 300. At this time, the setting unit 108A acquires information S9 on a trained model that satisfies the user's request more than the trained model set by the information S5 among trained models whose similarity is equal to or higher than a reference level. The setting unit 108A transmits the acquired information S9 to the electrostatic separation device 200A as operating condition information S10. The billing processing unit 110 generates billing information S11 based on the setting of the operating conditions in the electrostatic separation device 200A by the setting unit 108A, and transmits the generated billing information S11 to the user terminal device 500.

[0067] The acquisition unit 102 acquires a collection result S12 resulting from the operation of the electrostatic separation device 200A based on the operating condition information S10. The billing processing unit 110 detects, based on the collection result S12, that at least one of the collection rate, purity, environmental load, and operating cost has changed due to the operation of the electrostatic separation device 200A using operating conditions that satisfy the requirements set by the setting unit 108A. The billing processing unit 110 generates billing information based on the change in at least one of the collection rate, purity, environmental load, and operating cost due to the operation of the electrostatic separation device 200A using operating conditions that satisfy the requirements set by the setting unit 108A, and transmits the billing information to the user terminal device 500.

[0068] As described above, according to the setting support system 1A of the second embodiment, it is possible to set operating conditions for the electrostatic separation device 200A based on the user's request, the similarity, and the operating condition information of the electrostatic separation device 200B. In this way, according to the setting support system 1A, it is possible to support setting appropriate operating conditions for the electrostatic separation device 200A in response to the user's request.

[0069] According to the setting support system 1A of the second embodiment, it is possible to generate billing information based on setting the operating conditions that satisfy the user's request in the electrostatic separation device 200A. According to the setting support system 1A of the second embodiment, it is possible to generate billing information based on the collection result after the operating conditions based on the user's request are set in the electrostatic separation device 200A. As a result, according to the setting support system 1A, it is possible to build a business model in which the collection rate, purity, environmental load, operating cost, etc. of the electrostatic separation device 200A are adjusted to satisfy the user's request, thereby charging the user.

[0070] Although each embodiment and each variant have been described, these are merely examples and are not intended to be limiting. For example, any of the embodiments or variants, or a part of each embodiment or a part of each variant, may be combined with one or more other embodiments or one or more other variants to realize one aspect of the present invention. [Explanation of symbols]

[0071] 1, 1A... setting support system, 100, 100A... setting support device, 102... acquisition unit, 104... reception unit, 106... similarity calculation unit, 108, 108A... setting unit, 110... billing processing unit, 200, 200A, 200B... electrostatic separation device, 202... detection unit, 210... mixture, 220... supply unit, 221... hopper, 222... input feeder, 230... charging unit, 231... charging cylinder, 232... vibration feeder, 240... electric field generation unit, 241, 242... electrode, 243... direct current Power supply, 250... collection box, 251... housing section, 252, 253... partition plate, 254, 255... drive section, 261, 263... data processing section, 262... input amount conversion section, 264, 265, 266, 267... image processing section, 270... input plastic sensor section, 280... collected plastic sensor section, 300... database device, 310... user database, 320, 402... trained model, 400... learning device, 410... control device, 500... user terminal device

Claims

1. an acquisition unit that acquires mixture information related to a mixture including a plurality of types of objects in a first sorting device that sorts the mixture into groups of objects by type, and operating condition information of the first sorting device from a plurality of first sorting devices; a receiving unit that receives mixture information of the mixture to be sorted by the second sorting device; a similarity calculation unit that calculates a similarity between each of the mixture information accepted by the accepting unit and the mixture information acquired by the acquiring unit; a setting unit that selects operating condition information for setting operating conditions for the second sorting device from among the operating condition information of the plurality of first sorting devices based on the similarity calculated by the similarity calculation unit; and A setting support system comprising:

2. The setting support system according to claim 1 , wherein the mixture information includes at least one of a type, a composition, and a grain size of the object.

3. the sorting device is an electrostatic sorting device, The setting support system according to claim 1 , wherein the operating condition information is a magnitude of a voltage applied to an electrode pair that applies an electrostatic force to the charged object, or a position of a partition that divides an area that contains the object induced by the electrostatic force.

4. Each of the plurality of first sorting devices is trained on the mixture information, operating conditions, and recovery results as learning data, and performs sorting using a trained model trained to output operating conditions that result in optimal recovery results when mixture information is input, The trained model is created for each of the mixture information, The setting unit selects a trained model for setting operating conditions for the second sorting device from among a plurality of trained models of the plurality of first sorting devices. The setting support system according to claim 1 .

5. The setting support system described in claim 4, wherein the recovery results include a recovery rate of a specific object contained in the mixture and a purity of a specific object contained in the mixture.

6. The setting unit is Selecting a plurality of trained models based on the similarity; The setting support system of claim 4, further comprising: selecting operating condition information for setting operating conditions for the second selection device based on a single trained model obtained by combining the selected trained models.

7. a detection unit that detects mixture information of the mixture input to the second sorting device, the similarity calculation unit calculates a similarity between the mixture information detected by the detection unit and each of the mixture information acquired by the acquisition unit; the setting unit selects operating condition information for changing the operating conditions set in the second sorting device from the operating condition information of the plurality of first sorting devices based on the similarity calculated by the similarity calculation unit. The setting support system according to claim 1 .

8. The reception unit receives a request from a customer who owns the second sorting device, the setting unit selects operating condition information for setting operating conditions for the second sorting device from among the operating condition information of the plurality of first sorting devices based on a similarity between the request, each of the mixture information accepted by the accepting unit, and the mixture information acquired by the acquiring unit. The setting support system according to claim 1 .

9. The setting support system of claim 8, wherein the setting unit selects operating conditions that prioritize at least one of the recovery rate of a specific type of object in the mixture, the purity of a specific type of object in the mixture, the environmental load of the second sorting device, and the operating cost of the second sorting device based on the request.

10. 9. The setting support system according to claim 8, further comprising a billing processing unit that generates billing information based on the setting of the operating conditions that satisfy the request by the setting unit.

11. The setting support system described in claim 9, further comprising a billing processing unit that generates billing information based on a change in at least one of the recovery rate, the purity, the environmental load, and the operating cost due to the second sorting device operating under operating conditions that satisfy the requirements set by the setting unit.

12. The setting support system according to claim 8 , further comprising a billing processing unit that generates billing information when changed operating conditions are provided in response to a change in the mixture information or a change in the request being accepted by the accepting unit.

13. An acquisition unit that acquires mixture information regarding a mixture including a plurality of types of objects and a trained model that is created for each of the mixture information and outputs an operating condition when the mixture information is input; a reception unit that receives mixture information of the mixture to be sorted by the sorting device; a similarity calculation unit that calculates a similarity between the mixture information accepted by the acceptance unit and the mixture information acquired by the acquisition unit; A setting unit that selects a trained model for setting operating conditions of a selection device from among the plurality of trained models based on the similarity calculated by the similarity calculation unit; A setting support system comprising:

14. A step of acquiring mixture information related to a mixture including a plurality of types of objects in a first sorting device that sorts the mixture into each type of object and operating condition information of the first sorting device from a plurality of first sorting devices; receiving mixture information of the mixture to be sorted by the second sorting device; Calculating a similarity between the mixture information of the mixture sorted by the second sorting device and each of the mixture information regarding the mixture in the first sorting device; selecting operating condition information for setting operating conditions for the second sorting device from among the operating condition information of the plurality of first sorting devices based on the similarity; A setting assistance method comprising:

15. a similarity calculation unit that calculates a similarity between each of a plurality of mixture information related to a mixture in a plurality of first sorting devices that sort a mixture including a plurality of types of objects for each type of the objects and mixture information of a mixture sorted by a second sorting device; a setting unit that selects operating condition information for setting operating conditions for the second sorting device from among the operating condition information of the plurality of first sorting devices based on the similarity calculated by the similarity calculation unit; and A setting assistance device comprising:

16. The computer of the setting support device A step of calculating a similarity between each of a plurality of mixture information related to a mixture in a plurality of first sorting devices that sort a mixture including a plurality of types of objects for each type of the objects and mixture information of a mixture sorted by a second sorting device; selecting operating condition information for setting operating conditions for the second sorting device from among the operating condition information of the plurality of first sorting devices based on the similarity; A program to execute.

Images