Sorting processing system

Abstract

A sorting processing system according to the present disclosure comprises: a first detection unit that acquires first raw material information; a second detection unit that acquires second raw material information; a calculation unit that includes an analysis model for outputting a setting value related to a sorting condition; and a control unit that changes a sorting condition for a sorting device. In the calculation unit, the first raw material information is processed for each first processing cycle, the second raw material information is processed for each second processing cycle that is longer than the first processing cycle, and switching is performed between a first analysis method that includes outputting the setting value by inputting the first raw material information into the analysis model, and a second analysis method that includes outputting the setting value by inputting at least the second raw material information into the analysis model.

Description

Sorting and Processing System

[0001] The present disclosure relates to a sorting system.

[0002] Patent Document 1 discloses a technology for sorting mixtures of plastic pieces, etc. In such sorting, the state of the raw materials, such as the raw material composition ratio or specific charge, is detected, and sorting conditions are determined based on the results.

[0003] International Publication No. 2012 / 101874

[0004] The state of the raw material may change over time. To accommodate such changes, the state of the raw material is acquired at regular time intervals. In a configuration that detects multiple types of raw material states, the period for acquiring data required to determine the sorting conditions may differ for each raw material state. In this case, the accuracy of the sorting depends on how the multiple types of raw material states acquired at different periods are used to determine the sorting conditions.

[0005] In view of the above circumstances, the present disclosure aims to provide a sorting processing system that can ensure sorting accuracy in a configuration that acquires multiple types of raw material states at different cycles.

[0006] One aspect of the sorting processing system according to the present disclosure includes a sorting device that sorts a mixture containing multiple types of objects by object type; a first detection unit that acquires first raw material information regarding the mixture fed into the sorting device; a second detection unit that acquires second raw material information regarding the mixture fed into the sorting device, the second raw material information being of a different type from the first raw material information; a calculation unit having an analytical model that outputs setting values ​​regarding the sorting conditions of the sorting device; and a control unit that changes the sorting conditions of the sorting device based on the setting values ​​output by the calculation unit, wherein the calculation unit processes the first raw material information every first processing cycle and processes the second raw material information every second processing cycle that is longer than the first processing cycle, and the calculation unit switches between a first analysis method in which the first raw material information is input into the analytical model and the setting values ​​are output, and a second analysis method in which at least the second raw material information is input into the analytical model and the setting values ​​are output.

[0007] According to the sorting processing system of the present disclosure, it is possible to ensure sorting accuracy in a configuration in which multiple types of raw material states are acquired at different cycles.

[0008] Fig. 1 is a diagram showing a configuration example of a sorting processing system in embodiment 1. Fig. 2 is a diagram showing configuration examples of a first detection unit and a second detection unit of Fig. 1. Fig. 3 is a diagram explaining a method of acquiring a raw material composition ratio using HSI data of a hyperspectral camera. Fig. 4 is a diagram explaining a first analysis method and a second analysis method.

[0009] Embodiment 1. Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. Note that the scope of the present disclosure is not limited to the following embodiments and can be modified as desired within the scope of the technical concept of the present disclosure. FIG. 1 is a configuration diagram illustrating a sorting processing system 1 in embodiment 1. The sorting processing system 1 has a first detection unit 11, a sorting device 12, a calculation unit 13, a control unit 14, and a second detection unit 21.

[0010] The sorting processing system 1 is configured to sort a mixture containing multiple types of objects by object type. In this embodiment, a plastic piece group P will be described as an example of a "mixture containing multiple types of objects." In the example of FIG. 1, the plastic piece group P contains two types of plastic pieces p1 and p2 made of different materials. In this specification, the plastic pieces p1 and p2 may be referred to as "flakes" without distinguishing between the materials. Three or more types of plastic pieces may be included in the plastic piece group P.

[0011] In the following, an example will be described in which the plastic piece p1 is ABS and the plastic piece p2 is PS. The plastic pieces p1 and p2 are obtained, for example, by crushing the housing of a home appliance using a crusher and drying it. The plastic pieces p1 and p2 are formed to a size of, for example, about 10 mm square.

[0012] The sorting device 12 sorts the plastic pieces p1 and p2 by type using electrostatic sorting. In the example of Fig. 1, the sorting device 12 includes an input section 121, a charging cylinder 122, a vibrating feeder 123, a first electrode 124, a second electrode 125, a DC power supply 126, a collection box 127, and partition plates 128 and 129. However, the configuration of the sorting device 12 in Fig. 1 is merely an example and can be modified.

[0013] The sorting device 12 can electrostatically sort a group of plastic pieces P, which is a mixture of multiple types of plastic pieces p1 and p2 having different charging characteristics, into plastic pieces p1 and plastic pieces p2.

[0014] The input section 121 includes a hopper 121a and an input feeder 121b. Dried plastic piece groups P are supplied to the hopper 121a. The hopper 121a supplies a predetermined amount of the plastic piece groups P per unit time to the input feeder 121b. The input feeder 121b supplies the plastic piece groups P input from the hopper 121a into the charging cylinder 122.

[0015] The charging tube 122 and the vibrating feeder 123 constitute a charging unit. The charging unit charges each of the plastic pieces p1 and p2 and then drops them. Specifically, the charging tube 122 agitates the plastic piece group P by rotating. Inside the charging tube 122, the multiple types of plastic pieces p1 and p2 mixed in the plastic piece group P rub against each other and become charged. Each of the charged plastic pieces p1 and p2 has a charge of a polarity (positive or negative) according to the triboelectric series. In this example, the plastic piece p1, which is ABS, is positively charged, and the plastic piece p2, which is PS, is negatively charged.

[0016] The charged plastic pieces p1 and p2 are supplied to the rear end of the upper surface of the vibrating feeder 123. The positively charged plastic piece p1 and the negatively charged plastic piece p2 are attracted to each other by electrostatic force and become paired. The vibrating feeder 123 pushes the plastic pieces p1 and p2 forward while vibrating them up and down. This breaks the pairing of the plastic pieces p1 and p2, and the plastic pieces p1 and p2 move forward in the X direction in the figure. The plastic pieces p1 and p2 also fall from the vibrating feeder 123.

[0017] The electrodes 124 and 125 and the DC power supply 126 constitute an electric field generator. The electric field generator applies an electrostatic field to each charged plastic piece, causing each plastic piece to fall to a position corresponding to the charge state of the plastic piece. Specifically, the electrodes 124 and 125 are formed in a flat plate shape. The electrodes 124 and 125 are arranged in the X direction in the figure, facing each other and sandwiching the path along which the plastic pieces p1 and p2 fall. A ground voltage GND is applied to the first electrode 124. The DC power supply 126 applies a predetermined DC voltage between the first electrode 124 and the second electrode 125, generating an electrostatic field between the first electrode 124 and the second electrode 125.

[0018] When plastic pieces p1 and p2, which have been unpaired by the vibrating feeder 123, are dropped between the electrodes 124 and 125, each plastic piece falls while being attracted to either the electrode 124 side or the electrode 125 side by electrostatic force according to its charge state (polarity, amount of charge). In other words, each plastic piece p1 and p2 follows a parabolic trajectory according to its charge state and falls to a different position. In this example, plastic piece p1 is positively charged and therefore falls toward the first electrode 124. On the other hand, plastic piece p2 is negatively charged and therefore falls toward the second electrode 125.

[0019] The collection box 127 is provided below the electrodes 124 and 125 and collects the plastic pieces p1 and p2 that have fallen from the vibrating feeder 123 after passing between the electrodes 124 and 125. The collection box 127 is formed in a rectangular parallelepiped shape with an opening at the top. The opening of the collection box 127 is formed in a rectangular shape with its long side facing in the X direction in the figure.

[0020] Each of the partition plates 128, 129 is also referred to as a partition member. The partition plates 128, 129 are arranged parallel to the YZ plane in the figure within the recovery box 127 and are provided so as to be movable in the X direction in the figure. The positions of the partition plates 128, 129 in the X direction are controlled by the control unit 14. The partition plate 128 is located on the first electrode 124 side, and the partition plate 129 is located on the second electrode 125 side. The recovery box 127 is divided by the partition plates 128, 129 into a recovery chamber 127a on the first electrode 124 side, a recovery chamber 127b on the second electrode 125 side, and an intermediate recovery chamber 127c.

[0021] Each plastic piece p1, p2 that falls from the vibrating feeder 123 and passes between the electrodes 124, 125 is collected in one of three collection chambers 127a to 127c depending on its charge state. In this example, plastic piece p1 is positively charged and is collected in collection chamber 127a. On the other hand, plastic piece p2 is negatively charged and is collected in collection chamber 127b. Plastic pieces p1, p2 that are not sufficiently charged are collected in collection chamber 127c.

[0022] The plastic pieces p1 collected in recovery chamber 127a, the plastic pieces p2 collected in recovery chamber 127b, and the plastic pieces p1 and p2 collected in recovery chamber 127c are transported by a conveying machine (not shown) and stored in separate containers.

[0023] The first detection unit 11 acquires first raw material information. The first raw material information is, for example, a raw material composition ratio. When acquiring the raw material composition ratio, the first detection unit 11 includes, for example, a hyperspectral camera 11a and a processing unit 11b as shown in FIG. 2 . The hyperspectral camera 11a captures images of the plastic piece group P flowing on the vibrating feeder 123 using a line scan method. The hyperspectral camera 11a performs line scans on the plastic piece group P transported by the vibrating feeder 123 at a constant speed, for example, at approximately 500 fps.

[0024] The hyperspectral camera 11a separates light emitted from each point of the group of plastic pieces P, which is the subject, and captures it on the sensor surface. Multiple pixels are arranged on the sensor surface, and each pixel captures light of a different wavelength. The hyperspectral camera 11a outputs HSI data based on the intensity of light at each wavelength. HSI stands for hyperspectral imaging. The HSI data contains spectral information in the near-infrared range for each pixel.

[0025] 3 shows an example of acquiring a raw material composition ratio using HSI data from the hyperspectral camera 11a. As shown in FIG. 3(A), the HSI data includes spectral information for each pixel associated with the shape of the flakes on the vibrating feeder 123. Points (i) and (ii) are parts of the flakes, and point (iii) is part of the vibrating feeder 123. FIG. 3(A1) shows the spectra for each pixel at points (i) to (iii). In each graph in FIG. 3(A1), for example, the horizontal axis represents wavelength, and the vertical axis represents intensity for each wavelength.

[0026] The processing unit 11b of the first detection unit 11 processes the HSI data generated by the hyperspectral camera 11a. The processing unit 11b analyzes the HSI data and generates an image in which the flakes contained in the plastic fragment group P are color-coded according to the plastic material. FIG. 3(B) is an example of an image generated by analyzing the HSI data of FIG. 3(A). In the image of FIG. 3(B), flakes containing point (i) are displayed in red, and flakes containing point (ii) are displayed in blue. In the image of FIG. 3(B), for example, the red flakes are plastic fragments p1 and the blue flakes are plastic fragments p2. The processing unit 11b further performs image analysis to determine the area ratio of each color and convert it into a raw material composition ratio. In the example of FIG. 3(B), the ratio of the area of ​​the red region to the area of ​​the blue region is 60:40. Therefore, it can be determined that the raw material composition ratio of plastic fragment p1 is 60% and the raw material composition ratio of plastic fragment p2 is 40%. In this way, the first detection unit 11 sequentially measures the raw material composition ratio of the group of plastic pieces P conveyed by the vibrating feeder 123 using the line scan method.

[0027] The second detection unit 21 acquires second raw material information, which is different from the first raw material information. The second raw material information is, for example, the specific charge. To acquire the specific charge, the second detection unit 21 includes, for example, a charge amount sensor 21a and a weight sensor 21b as shown in FIG. 2. Specifically, the second detection unit 21 picks up flakes transported on the vibrating feeder 123 one by one, and measures the charge amount and weight using the charge amount sensor 21a and the weight sensor 21b. The specific charge of each flake can be measured by dividing the charge amount by the weight.

[0028] The first detection unit 11 inputs the acquired first raw material information to the calculation unit 13. The second detection unit 21 inputs the acquired second raw material information to the calculation unit 13. The calculation unit 13 is equipped with an analytical model that outputs setting values ​​for the sorting conditions of the sorting device 12. This analytical model analyzes the setting values ​​of the sorting conditions based on at least one of the first raw material information and the second raw material information so as to obtain optimal sorting accuracy. The analytical model of the calculation unit 13 may use, for example, multiple regression analysis. Alternatively, the analytical model of the calculation unit 13 may use machine learning. These analytical models perform analysis based on a combination of past raw material information, setting values ​​of the sorting conditions, and sorting accuracy, and output setting values ​​of the sorting conditions that are appropriate for the current raw material information.

[0029] The set values ​​of the sorting conditions, which are the analysis results of the calculation unit 13, are input to the control unit 14. The control unit 14 controls the sorting conditions of the sorting device 12 based on the set values ​​input from the calculation unit 13. For example, the control unit 14 may control the positions of the partition plates 128 and 129 in the X direction. Alternatively, the control unit 14 may control the rotation speed or tilt of the charging cylinder 122, the voltage applied by the DC power supply 126 to the second electrode 125, etc. In other words, the "set values" are the positions of the partition plates 128 and 129 in the X direction, the rotation speed or tilt of the charging cylinder 122, the voltage of the DC power supply 126, etc. The "set values" may include these multiple parameters.

[0030] Here, the calculation unit 13 processes the first ingredient information acquired by the first detection unit 11 based on a "first processing cycle." The calculation unit 13 also processes the second ingredient information acquired by the second detection unit 21 based on a "second processing cycle." The first processing cycle and the second processing cycle are determined based on the amount of data required for the first ingredient information and the second ingredient information. The "required data amount" refers to the amount of data for the first ingredient information and the second ingredient information required by the calculation unit 13 to analyze the set values ​​of the sorting conditions. The required data amount is determined by the variability in the measurement data of the first ingredient information and the second ingredient information. The greater the variability in the measurement data, the greater the required data amount is to ensure sorting accuracy.

[0031] As an example, when calculating the set values ​​of the sorting conditions based on the raw material composition ratio, a data volume of 500 flakes is required to ensure sorting accuracy. With the line scanning method, for example, the raw material composition ratio of 1,000 flakes can be obtained per minute. In this example, the required data volume, i.e., the "raw material composition ratio of 500 flakes," is obtained in a period of 30 seconds. Therefore, the "first processing period" in this example is 30 seconds.

[0032] As an example, when calculating the set value of the sorting condition based on the specific charge, a data volume of 200 flakes is required to ensure sorting accuracy. In the picking method, the specific charge of 10 flakes can be obtained, for example, per minute. In this example, the required data volume, i.e., the "specific charge of 200 flakes," is obtained over a period of 20 minutes. Therefore, the "second processing period" in this example is 20 minutes.

[0033] As described above, the first raw material information and the second raw material information include measurement data for multiple flakes. The calculation unit 13 statistically processes this measurement data to obtain statistical values ​​such as the average or median. The calculation unit 13 inputs these statistical values ​​into an analysis model to output setting values ​​for the sorting conditions. Note that the specific values ​​for the first processing period, second processing period, requested data volume, etc. described above are merely examples and can be changed. However, in the present disclosure, the first processing period is shorter than the second processing period.

[0034] Here, the first raw material information has the advantage that, because the first processing cycle is short, any fluctuation points in the mixture being input are easily detected. A "fluctuation point" refers to, for example, a change in the type of home appliance that formed the original mixture input into the sorting device 12. It is expected that the first raw material information and the second raw material information will fluctuate significantly around the fluctuation point. Due to the long second processing cycle, the second raw material information is likely to include data from before and after the fluctuation point. If the setting values ​​for the sorting conditions are determined based on the second raw material information that includes data from before and after the fluctuation point, sorting accuracy is likely to decrease.

[0035] On the other hand, the raw material composition ratio, which is an example of the first raw material information, has a low correlation with sorting accuracy such as recovery rate and purity. The specific charge, which is an example of the second raw material information, has a high correlation with sorting accuracy compared to the raw material composition ratio. Therefore, by determining the setting values ​​of the sorting conditions using the second raw material information as much as possible, sorting accuracy can be further improved.

[0036] In consideration of the above circumstances, the calculation unit 13 in this embodiment switches the analysis method when outputting the setting values ​​of the sorting conditions, as shown in Figure 4. In the "first analysis method" shown in Figure 4(b), the calculation unit 13 inputs the first raw material information into the analysis model without using the second raw material information, and outputs the setting values ​​of the sorting conditions. In the "second analysis method" shown in Figure 4(c), the calculation unit 13 inputs both the first raw material information and the second raw material information into the analysis model, and outputs the setting values ​​of the sorting conditions.

[0037] FIG. 4(a) is an example of a graph showing fluctuations in the first and second raw material information in chronological order. In FIG. 4(a), the plots of the first and second raw material information indicate statistical values ​​for the required data amount calculated for each processing cycle. In this example, the "statistical values" are average values, but medians, etc., may be used instead. In FIG. 4(a), raw material composition ratios are illustrated as the first raw material information, and specific charge is illustrated as the second raw material information. At time t1, statistical values ​​for the required data amount of the first raw material information have been acquired, but statistical values ​​for the required data amount of the second raw material information have not been acquired. Therefore, between times t1 and t2, the calculation unit 13 outputs the setting values ​​of the sorting conditions using the "first analysis method."

[0038] At time t2 in Figure 4(a), statistical values ​​of the second ingredient information for the required data amount are acquired. Therefore, from time t2 onwards, it is highly likely that more preferable sorting accuracy will be obtained by outputting the setting values ​​of the sorting conditions using both the first ingredient information and the second ingredient information. Therefore, between times t2 and t3, the calculation unit 13 outputs the setting values ​​of the sorting conditions using the "second analysis method."

[0039] At time t3 in FIG. 4( a), the value of the first ingredient information fluctuates significantly. Therefore, the calculation unit 13 can determine that time t3 is a "fluctuation point." For example, the calculation unit 13 may compare the current statistical value of the first ingredient information with the previous statistical value, and determine that the "fluctuation point" exists when the difference exceeds a threshold. It is expected that not only the first ingredient information but also the second ingredient information fluctuates significantly around time t3, which is the fluctuation point. Therefore, if statistical values ​​including the second ingredient information acquired before time t3 are used after time t3, it is highly likely that desirable sorting accuracy will not be achieved. Therefore, the calculation unit 13 outputs the setting values ​​of the sorting conditions using the "first analysis method" between times t3 and t4.

[0040] At time t4 in Figure 4(a), statistical values ​​of the second ingredient information for the required data amount are acquired. However, the statistical values ​​of the second ingredient information acquired at time t4 include the state of the mixture prior to the fluctuation point. Therefore, the statistical values ​​of the second ingredient information at time t4 may not accurately reflect the current state of the ingredients. Therefore, the calculation unit 13 continues to use the "first analysis method" between times t4 and t5.

[0041] At time t5 in FIG. 4(a), the statistical value of the second ingredient information is obtained. At time t5, at least one second processing cycle has passed since time t3, which is the fluctuation point. Therefore, the statistical value of the second ingredient information obtained at time t5 does not include the state of the mixture prior to the fluctuation point. In other words, the statistical value of the second ingredient information at time t4 has low data reliability, while the statistical value of the second ingredient information at time t5 has high data reliability. Therefore, from time t5 onwards, the calculation unit 13 outputs the setting values ​​of the sorting conditions using the "second analysis method."

[0042] Thereafter, the calculation unit 13 continues to detect the presence or absence of a fluctuation point based on the first ingredient information. If a fluctuation point is detected, the first analysis method is used until at least one second processing cycle has elapsed. After one second processing cycle has elapsed after the fluctuation point, the calculation unit 13 uses the second analysis method. This allows the calculation unit 13 to output the setting values ​​of the sorting conditions with high accuracy using both the first ingredient information and the second ingredient information, while excluding the use of the second ingredient information, which contains a mixture of data before and after the fluctuation point.

[0043] In the second analysis method, if good and appropriate setting values ​​for the sorting conditions can be output using only the second raw material information, the first raw material information does not need to be used. In other words, it is sufficient to use at least the second raw material information in the second analysis method.

[0044] As described above, the sorting processing system 1 according to the present disclosure includes a sorting device 12 that sorts a mixture containing multiple types of objects by object type, a first detection unit 11 that acquires first raw material information regarding the mixture input to the sorting device 12, a second detection unit 21 that acquires second raw material information regarding the mixture input to the sorting device 12, the second raw material information being different from the first raw material information, a calculation unit 13 that includes an analytical model that outputs setting values ​​related to the sorting conditions, and a control unit 14 that changes the sorting conditions of the sorting device 12 based on the setting values ​​output by the calculation unit 13. The calculation unit 13 processes the first raw material information every first processing cycle and processes the second raw material information every second processing cycle that is longer than the first processing cycle. The calculation unit 13 switches between a first analysis method in which the first raw material information is input to the analytical model and a second analysis method in which at least the second raw material information is input to the analytical model and a setting value is output. According to this sorting processing system 1, it is possible to ensure sorting accuracy in a configuration in which statistical values ​​of the first raw material information and the second raw material information are obtained in a first processing cycle and a second processing cycle that are different from each other.

[0045] Furthermore, the calculation unit 13 may detect a fluctuation point in the state of the mixture input to the sorting device 12 based on the first raw material information processed in each first processing cycle, and output a set value using the first analysis method after the fluctuation point until at least the second processing cycle has elapsed. This configuration prevents the set value of the sorting condition from being determined using the second raw material information that includes data before the fluctuation point. Therefore, it is possible to prevent a decrease in sorting accuracy at the fluctuation point.

[0046] Furthermore, the first raw material information may be the raw material composition ratio for each material in the plastic piece group P, and the first detection unit 11 may have a hyperspectral camera 11a that captures the plastic piece group P. In a configuration in which the hyperspectral camera 11a is used to detect the raw material composition ratio, it is possible to obtain the raw material composition ratio for 1,000 flakes per minute, for example. Therefore, the raw material composition ratio can be suitably used as the first raw material information for detecting a fluctuation point.

[0047] The technical scope of the present disclosure is not limited to the above-described embodiments, and various modifications can be made without departing from the spirit of the present disclosure.

[0048] For example, in the above embodiment, a group of plastic pieces P was described as an example of a "mixture containing multiple types of objects" to be sorted by the sorting system 1. However, the sorting system 1 may also sort a mixture of objects other than plastic pieces. In this case, too, the present disclosure contributes to solving the problem of ensuring sorting accuracy in a configuration in which first raw material information and second raw material information are acquired in different first and second processing cycles.

[0049] In the above embodiment, statistical values ​​(average values, etc.) of the first ingredient information and the like are input to the analysis model. However, calculating statistical values ​​is not essential. For example, the detection data from the first detection unit 11 or the second detection unit 21 may be directly input to the analysis model, and the setting values ​​for the sorting conditions may be output. In other words, "processing the first ingredient information for each first processing cycle" includes not only calculating statistical values ​​but also inputting the required amount of first ingredient information into the analysis model.

[0050] In the above embodiment, the raw material composition ratio is exemplified as the first raw material information, and the specific charge is exemplified as the second raw material information. As other examples, the first raw material information may be the input amount of the plastic piece group P, or the temperature and humidity in the environment of the sorting device 12. The second raw material information may be the moisture content in the flakes. Other types of first raw material information or second raw material information may also be used. The calculation unit 13 may output setting values ​​for the sorting conditions based on three or more pieces of raw material information.

[0051] That is, as third raw material information different from the first raw material information and the second raw material information, the input amount of the plastic piece group P, the temperature and humidity in the environment of the sorting device 12, the moisture content in the flakes, etc. may be used. In this case, in the first analysis method in FIG. 4(b), the first raw material information and the third raw material information may be input to the analysis model. In the second analysis method in FIG. 4(c), the first to third raw material information, or only the second raw material information and the third raw material information, may be input to the analysis model.

[0052] Furthermore, each function of the calculation unit 13, the control unit 14, etc. is realized by a processor such as a CPU (Central Processing Unit) executing a program stored in a program memory. Some or all of these functions may be realized by hardware such as an LSI (Large Scale Integration), an ASIC (Application Specific Integrated Circuit), or an FPGA (Field-Programmable Gate Array), or may be realized by a combination of software and hardware. The functions of the calculation unit 13 and the control unit 14 may be realized by the same hardware.

[0053] In the above embodiment, the case where plastic pieces are sorted by electrostatic sorting has been described. However, the sorting method is not limited to this, and may be, for example, gravity sorting or optical sorting. Gravity sorting is a sorting method that utilizes the fact that each type of plastic piece has a different specific gravity. For example, if a group of plastic pieces is vibrated or floated on a medium, plastic pieces with a higher specific gravity will descend and plastic pieces with a lower specific gravity will rise.

[0054] Optical sorting is a sorting method that takes advantage of the fact that the reflectivity of light differs depending on the type of plastic piece. In optical sorting, a group of plastic pieces is irradiated with detection light and the reflected light is detected. Light of various wavelength bands, such as infrared or X-ray, can be used as the detection light. By detecting the spectrum of reflected light or Raman scattered light, the plastic pieces can be distinguished by type. After distinguishing them in this way, the plastic pieces can be sorted by air blowing, etc.

[0055] In addition, the above-described embodiments and modifications may be combined as appropriate.

[0056] REFERENCE SIGNS LIST 1... Sorting processing system 11... First detection unit 11a... Hyperspectral camera 12... Sorting device 13... Calculation unit 14... Control unit 21... Second detection unit P... Plastic piece group

Claims

1. A sorting processing system comprising: a sorting device that sorts a mixture containing multiple types of objects by type of object; a first detection unit that acquires first raw material information regarding the mixture fed into the sorting device; a second detection unit that acquires second raw material information regarding the mixture fed into the sorting device, the second raw material information being of a different type from the first raw material information; a calculation unit having an analytical model that outputs setting values ​​regarding the sorting conditions of the sorting device; and a control unit that changes the sorting conditions of the sorting device based on the setting values ​​output by the calculation unit, wherein the calculation unit processes the first raw material information for each first processing cycle and processes the second raw material information for each second processing cycle that is longer than the first processing cycle, and the calculation unit switches between a first analysis method in which the first raw material information is input into the analytical model and the setting values ​​are output, and a second analysis method in which at least the second raw material information is input into the analytical model and the setting values ​​are output.

2. The sorting processing system described in claim 1, wherein the calculation unit detects a fluctuation point in the state of the mixture fed into the sorting device based on the first raw material information processed for each first processing cycle, and outputs the set value using the first analysis method after the fluctuation point until at least the second processing cycle has elapsed.

3. The sorting system according to claim 1 or 2, wherein the sorting device sorts the mixture by electrostatic sorting.

4. The sorting processing system according to any one of claims 1 to 3, wherein the mixture is a group of plastic pieces and the type of object is a plastic material.

5. The sorting processing system described in claim 4, wherein the first raw material information is the raw material composition ratio for each material in the group of plastic pieces, and the first detection unit has a hyperspectral camera that images the group of plastic pieces.

Images