Resin identification method

Abstract

To provide a resin determination method for accurately determining resin in which carbon black is contained.SOLUTION: Based on reflected light 4 received from resin 2 of a selection target irradiated with infrared light 3, a reflection or absorption spectrum of the resin is calculated. Maximum intensity of the calculated spectrum is obtained, and when it is equal to or more than a first threshold value of maximum obtainable intensity, it is determined that carbon black is not contained and a resin type is determined after normalization. When the maximum intensity is less than the first threshold value, it is determined that the carbon black is contained as a pseudo-containment, and the resin type is determined.SELECTED DRAWING: Figure 5

Description

Technical Field

[0001] The present invention relates to a resin determination method for determining a resin type in a sorting target in which a plurality of types of small pieces are gathered without contact.

Background Art

[0002] Due to mass consumption and mass disposal-type economic activities, global environmental problems such as global warming or resource depletion are occurring.

[0003] Under such circumstances, in order to build a resource recycling-oriented society, in Japan, the Home Appliance Recycling Law has been implemented since April 2001. According to the Home Appliance Recycling Law, the recycling of used home appliances (such as air conditioners, televisions, refrigerators, freezers, washing machines, and clothes dryers) is mandatory. As a result, used home appliances are crushed into small pieces at a home appliance recycling factory and then sorted and recovered by material type using magnetism, wind power, vibration, etc., and recycled as recycled materials. In resin materials, polypropylene (hereinafter referred to as PP), polystyrene (hereinafter referred to as PS), and acrylonitrile-butadiene-styrene (hereinafter referred to as ABS) are widely used in home appliances, and as proposed in Patent Document 1, they are sorted and recovered by resin type using a sorting device that utilizes the light absorption characteristics in the near-infrared region (wavelength range 1 to 3 μm) due to the molecular structure of the resin.

[0004] This sorting device irradiates light including the near-infrared region onto small pieces conveyed by a conveyor, detects the reflection or absorption spectrum from the resin without contact, and can determine the resin type, so that a large number of small pieces can be sorted.

[0005] However, in black resin etc. containing carbon black (fine particles mainly composed of carbon) used in small home appliances or automobiles, etc., since the light in the near-infrared region is absorbed, it is a major problem that a sufficient reflection or absorption spectrum cannot be obtained and sorting cannot be performed.

[0006] Therefore, a device that takes into account the aforementioned problems regarding the recycling of resin materials has been proposed in Patent Document 1. In the technology described in Patent Document 1, a material identification device 101 for black waste plastics, as shown in Figure 10, is used to irradiate black waste plastic pieces 102 flowing on a conveyor belt 103 with an infrared light source 104, detect the reflected light with a mid-infrared sensor 105, and obtain a reflection spectrum in the mid-infrared region (wavelength range 3-5 μm) where the influence of carbon black is reduced. This makes it possible to identify not only white resins that do not contain carbon black, but also black resins, which were difficult to identify in the near-infrared region.

[0007] Furthermore, in recent years, black pigments that do not contain carbon black have been developed and are proposed in Patent Document 2. Unlike carbon black, which absorbs infrared rays, these black pigments have the property of reflecting infrared rays, making it possible to recover black resins that were previously difficult to separate in the near-infrared region. [Prior art documents] [Patent Documents]

[0008] [Patent Document 1] Patent No. 5367145 [Patent Document 2] Patent No. 5932464 [Overview of the Initiative] [Problems that the invention aims to solve]

[0009] In the future, as both black resins containing carbon black and black resins without carbon black become widespread in the market, it is anticipated that recycled resins collected from the market will contain a mixture of both types. In this case, if these different black pigments are not separated and collected properly and end up mixed together unintentionally, it will affect the color adjustment during recycling or the resin's functions, such as heat resistance. Therefore, it is necessary to accurately determine whether or not carbon black is present.

[0010] However, in the mid-infrared region, the presence of carbon black did not significantly alter the absorption or reflectance spectrum from the resin, making it difficult to determine whether carbon black was present. Furthermore, in the near-infrared region, while the presence of carbon black did significantly alter the absorption or reflectance spectrum, resins colored gray with less than 1% carbon black showed only slight attenuation of the absorption or reflectance spectrum, resulting in a spectrum very similar to that of white resins, making determination difficult.

[0011] The present invention aims to solve the above problems and to provide a resin determination method that accurately determines whether a resin contains carbon black by utilizing the attenuation effect of reflected light due to the inclusion of carbon black when determining the resin using the reflectance or absorption spectrum in the near-infrared region. [Means for solving the problem]

[0012] To achieve the above objective, the present invention is configured as follows.

[0013] According to one aspect of the present invention, Including the near-infrared region If the reflected or absorbed spectrum obtained from the reflected light received from the object to be sorted after being irradiated with infrared light is equal to or greater than a predetermined threshold (where the intensity of the reflected or absorbed spectrum is set as the first threshold), then it is normalized and obtained in advance. 2 The first similarity is calculated with the spectral data of more than one type of resin. The aforementioned first similarity is, Is it above the pre-set second threshold for the first similarity? Tsu High Spectral data of resin types that do not contain carbon black This is determined to be the resin type of the object to be sorted, If the reflection or absorption spectrum obtained by the reflected light is less than the first threshold, Two types of resins, including one containing carbon black and one without carbon black. We calculate the second similarity with the spectral data of more than one type of resin. The second similarity is, Is it above the pre-set third threshold for the second similarity? TsuAlso high Spectral data for resin types containing carbon black or resin types that do not contain carbon black is determined as the resin type of the object to be sorted.

Advantages of the Invention

[0014] As described above, according to the resin determination method according to the above aspect of the present invention, based on the reflected light received from the resin of the object to be sorted irradiated with infrared light, the reflection or absorption spectrum of the resin is calculated. Among the calculated spectra, the maximum intensity of the spectrum is obtained. If it is equal to or higher than the first threshold value of the maximum achievable intensity, it is determined that carbon black is not contained. After normalization, the resin type is determined. If it is less than the first threshold value, it is determined whether the resin contains carbon black as pseudo-contained, and the resin type is determined. As a result, when determining and sorting the resin using the reflection or absorption spectrum of the object, the content of carbon black can be accurately determined.

Brief Description of the Drawings

[0015] [Figure 1] Schematic diagram of the resin determination device in the embodiment of the present invention [Figure 2] Schematic diagram showing the influence of the mixing of carbon black-containing resin by conventional mid-infrared sorting [Figure 3] Schematic diagram showing the influence of the mixing of carbon black-containing resin by conventional near-infrared sorting [Figure 4] Graph of the spectrum in the conventional resin determination in the near-infrared region where the horizontal axis is wavelength and the vertical axis is absorbance [Figure 5] Flowchart showing the flow of the resin determination device determining the resin in the embodiment of the present invention [Figure 6] Graph of the spectrum of the resin sample in the near-infrared region where the horizontal axis is wavelength and the vertical axis is absorbance [Figure 7] Diagram showing the determination result of the conventional method [Figure 8] Diagram showing the determination result according to the embodiment of the present invention [Figure 9]A schematic diagram showing the effect of near-infrared sorting on the inclusion of carbon black-containing resin according to an embodiment of the present invention. [Figure 10] Schematic diagram of the conventional resin determination apparatus described in Patent Document 1 [Modes for carrying out the invention]

[0016] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

[0017] (Embodiment) Figure 1 is a schematic diagram of a resin determination device 1 capable of implementing the resin determination method according to the embodiment.

[0018] Resin 2 is a resin whose type is unknown, including black resin containing carbon black (i.e., carbon-based fine particles), black resin without carbon black, white resin, translucent resin, or gray resin. Therefore, Resin 2 refers to resin subject to sorting 2. Resin type refers to the resin type of the subject to sorting.

[0019] The configuration of the resin determination device 1, which accurately determines the type of resin based on whether or not carbon black is contained in resin 2, will be explained using Figure 1.

[0020] The resin determination device 1 comprises an infrared detection unit 8, a digital data conversion device 9, and a processing unit 10.

[0021] The infrared detection unit 8 has the function of irradiating the resin 2 with infrared light and the function of receiving the reflected light 4 from the resin 2 of the irradiated infrared light 3. Infrared light has absorption characteristics in the near-infrared region (effective wavelength range 1 to 3 μm).

[0022] The digital data conversion device 9 converts the electrical signal output by the infrared detection unit 8 in response to the reflected light 4 into digital data.

[0023] The processing unit 10 calculates the reflectance spectrum of the resin 2 based on the digital data output from the digital data conversion device 9.

[0024] In Figure 1, the belt conveyor 5 is moving at a constant speed and is an example of a transfer section for transporting resin 2. The resin 2 is transported along the longitudinal direction of the belt conveyor 5 from the input area 6 to the detection area 7.

[0025] Furthermore, an infrared detection unit 8 is positioned above the detection area 7 of the belt conveyor 5.

[0026] The arithmetic processing unit 10 analyzes the information output from the digital data conversion device 9 to obtain the reflectance spectrum of the resin 2 (i.e., reflectance spectrum acquisition function). The arithmetic processing unit 10 also performs normalization and other processing on the obtained reflectance spectrum to make it easier to determine (i.e., normalization processing function). Furthermore, the arithmetic processing unit 10 compares the detected reflectance spectrum data with the resin reflectance spectrum data that has been registered in advance (i.e., sample spectrum data) to determine the type of resin (i.e., resin type determination function).

[0027] Here, we will briefly explain how the processing unit 10 calculates the reflectance spectrum from the input digital data (i.e., the reflectance spectrum acquisition function). The electrical signal photoelectrically converted by the infrared detection unit 8 in response to the reflected light 4 depends on the intensity of the received light. Therefore, it is possible to obtain information on the intensity of the reflected light 4 from the resin 2 from the digital data converted by the digital data conversion device 9.

[0028] Before describing the resin type determination method (i.e., resin type determination function) according to the first embodiment, we will first explain the effect of contamination by resins containing carbon black when resins are sorted using a conventional determination method.

[0029] First, in conventional methods where sorting is performed in the mid-infrared region, the carbon black content does not affect the spectrum, so even resins of different colors will obtain the same spectrum if they are of the same resin type. For example, if resins of different colors such as white, translucent, gray (containing and not containing carbon black), or black (containing and not containing carbon black) are transported, as shown in Figure 2, they will all be sorted and recovered as the same resin type, resulting in all the carbon black-containing resins being mixed in with the carbon black-free resins.

[0030] Furthermore, in the case of sorting in the near-infrared region as in the conventional example, if resins of different colors as described above are transported, as shown in Figure 3, resins that do not contain carbon black do not experience spectral attenuation. Therefore, if gray and black resins without carbon black are the same type of resin, they will obtain the same spectrum and will be recovered as the same type of resin. Among the resins that do contain carbon black, black resins containing a few percent are greatly affected by attenuation and are therefore recovered as non-recoverable. On the other hand, gray resins containing less than 1% carbon black have a spectrum very similar to white resins, as shown in Figure 4, and are therefore sorted and recovered as the same type of resin, resulting in carbon black-containing resins being mixed in with carbon black-free resins.

[0031] Therefore, the inventors have found a method for accurately determining whether a resin contains carbon black, that is, a method for accurately determining whether the resin according to the first embodiment is present, by utilizing the effect of light attenuation due to the inclusion of carbon black.

[0032] Next, we will explain a resin determination method that takes into account the carbon black content using the resin determination device 1.

[0033] The infrared detection unit 8 shall include two or more light sources with a broad wavelength range, such as a blackbody light source or a halogen lamp, or single-wavelength light sources with an absorption wavelength range corresponding to the resin to be detected. The infrared detection unit 8 shall also include a light-receiving element that receives reflected light of each wavelength from the above-mentioned light sources.

[0034] The digital data conversion device 9 converts analog data from the light-receiving element into digital data and sends it to the arithmetic processing unit 10.

[0035] The processing unit 10 calculates the reflection or absorption intensity of each wavelength, i.e., the spectrum, based on the digital data output from the photodetector, and determines the type of resin by evaluating the spectrum.

[0036] The specific resin determination method will be explained in detail based on the determination flowchart shown in Figure 5. First, in step S1, the maximum intensity of the obtained spectrum is determined to be greater than or less than the first threshold, using an arbitrary percentage of the maximum obtainable intensity, for example, 55%, as the first threshold from the spectrum obtained from the arithmetic processing unit 10. If the maximum intensity of the obtained spectrum is greater than or equal to the first threshold, it is determined to be a resin that does not contain carbon black, i.e., a carbon black-free resin, and the process proceeds to step S2.

[0037] Next, in step S2, the spectrum is normalized.

[0038] Next, in step S3, the resin type is determined by Euclidean distance as an example of first similarity. That is, the first similarity is calculated between the normalized spectral data and the spectral data of one or more resin types that have been acquired in advance, and the one that is above a pre-set second threshold for first similarity and has the highest first similarity is determined to be the resin type of resin 2, for example, resin A and resin B.

[0039] On the other hand, if the maximum intensity of the spectrum obtained in step S1 is less than the first threshold, the resin is determined to potentially contain carbon black, i.e., a resin that pseudo-contains carbon black, and the process proceeds to step S4.

[0040] Next, in step S4, the resin containing carbon rack is determined by Euclidean distance as an example of second similarity. That is, the second similarity is calculated between the detected spectral data and the spectral data of one or more resin types that have been previously obtained. The resin with the highest second similarity that is above the pre-set third threshold for second similarity is determined to be resin type 2, for example, resin A or resin B that does not contain carbon black, and resin A' or resin B' that contains carbon black.

[0041] In steps S3 and S4, for resin determination, a sample spectrum is obtained from a resin whose physical properties are known in advance. The Euclidean distance d between the spectrum obtained from the resin to be determined and the sample spectrum is calculated using the following formula (1). The resin type is determined by a determination algorithm that compares this result with a pre-set second or third threshold for the Euclidean distance.

[0042] For example, the first similarity is calculated using the Euclidean distance for each resin, and the resin with the highest first similarity that is above the second threshold is determined to be the resin type of resin 2, for example, resin A and resin B.

[0043] The second similarity is calculated using the Euclidean distance for each, and the resin with the highest second similarity that is above the third threshold for the second similarity is determined to be resin type 2, for example, resin A or resin B, and resin A' or resin B'.

[0044] The second and third thresholds are set to thresholds that allow for correct determination based on the similarity between spectral data of one or more resin types that have been acquired in advance.

[0045]

number

[0046] Here, x i : Spectral intensity of the object to be judged, x' i :Sample spectral intensity, i:Variable on the horizontal axis (wavelength) as shown in Figure 4, etc.

[0047] Next, we will describe an experiment that confirmed the effectiveness of a method for accurately determining whether a resin contains carbon black according to the first embodiment.

[0048] As shown in Figure 6, the experiment was conducted using five samples of resin 2: white, translucent, gray (containing carbon black), black (containing carbon black), and black (not containing carbon black), and the spectra were determined. Figure 6 shows the results when PP was used as the resin matrix. The reason for excluding the gray (not containing carbon black) sample is that the results for the black (not containing carbon black) sample were so similar that no difference was observed, and the gray (not containing carbon black) sample was excluded as it would be redundant. Therefore, the results for the gray (not containing carbon black) sample are treated as the same as those for the black (not containing carbon black) sample shown below.

[0049] The Euclidean distance was determined using the resin spectrum obtained above, and the effectiveness of the conventional method and the embodiment of the present invention was verified.

[0050] For the sample spectrum in equation (1), the resin spectrum obtained in Figure 6 was used as the sample.

[0051] First, in the conventional method, the obtained spectrum is normalized, and the Euclidean distance is calculated using equation (1). The calculation results are shown in Figure 7. The value calculated by equation (1) approaches 0 as the degree of agreement with the sample spectrum increases. When looking at the values ​​for gray (containing carbon black), white is 0.40 and translucent is 0.45, showing close values. This is because the spectrum approximates after standardization, as shown in Figure 4. In this case, considering the spectral variability of the resin actually generated during recycling, there is a possibility of misidentification as white or translucent, making it difficult to determine whether the resin contains carbon black.

[0052] Next, Figure 8 shows the results calculated based on the judgment flowchart shown in Figure 5, using the resin determination method that takes into account the carbon black content of the embodiment of the present invention. The three colors that fell below the first threshold for maximum acquired intensity were translucent, gray (containing carbon black), and black (containing carbon black). Looking at the degree of agreement, or similarity, when targeting gray (containing carbon black), the translucent color was 0.63 and the black (containing carbon black) was 1.25. Compared with the results of the conventional method, the 40th percentile value was significantly higher for the translucent color, indicating a reduction in the risk of misidentification. Furthermore, regarding the risk of misidentification with white, since white is above the first threshold for maximum intensity and is excluded from the determination candidates, the risk of misidentification is eliminated. In addition, by setting the third threshold to 0.5 when determining whether a resin contains carbon rack after it has been determined to be a carbon black pseudo-containing resin, it becomes possible to determine without misidentification, and as shown in Figure 9, it is possible to accurately determine whether carbon black is contained and recover carbon black-free resin.

[0053] Here, the method using Euclidean distance as an example of similarity was explained as a judgment algorithm. However, other examples of similarity, such as regression analysis or correlation coefficients, may be selected as appropriate, and similarly, thresholds for strength judgment or resin judgment may be selected as appropriate.

[0054] Furthermore, although the above description used PP as the base resin, the resin determination method of the embodiment of the present invention addresses only the influence of carbon black content. Therefore, even with other resin base materials such as ABS or PS, the carbon black content can be determined with the same accuracy as with PP, and carbon black-free resin can be recovered.

[0055] As described above, the resin determination method according to the embodiment utilizes the attenuation effect of reflected light 4 due to the presence of carbon black when determining the resin 2 using the reflectance or absorption spectrum in the near-infrared region. Specifically, the reflectance or absorption spectrum of the resin 2 is calculated based on the reflected light 4 received from the resin 2 of the object to be sorted, which has been irradiated with infrared light 3. The maximum intensity of the spectrum is determined from the calculated spectrum, and if it is above a first threshold of the maximum obtainable intensity, it is determined that carbon black is not present. After normalization, the resin type is determined, and if it is below the first threshold, it is determined whether the resin contains carbon black as a pseudo-container, and the resin type is determined. As a result, resins containing carbon black can be determined with high accuracy.

[0056] Furthermore, by appropriately combining any embodiment or modification from the various embodiments or modifications described above, the effects of each can be achieved. In addition, it is possible to combine embodiments with each other, or embodiments with each other, or embodiments with each other, as well as to combine features from different embodiments or embodiments. [Industrial applicability]

[0057] As described above, the resin determination method according to the above aspect of the present invention can quickly determine whether or not carbon black (carbon-based fine particles) is present in black resins and the like that contain carbon black, and can therefore be used in recycling processes and the like that require the rapid sorting of multiple items. [Explanation of Symbols]

[0058] 1. Resin determination device 2 resin 3. Irradiated light 4 Reflected light 5 Belt conveyor 6 Input area 7 Detection area 8. Infrared detection unit 9. Digital Data Conversion Device 10 Arithmetic Processing Unit

Claims

1. If the reflection or absorption spectrum obtained from the reflected light received from the object to be sorted, which is irradiated with infrared light including the near-infrared region, is equal to or greater than a first threshold set by a preset, the intensity of the reflection or absorption spectrum is normalized to obtain a first similarity with spectral data of two or more types of resins that have been acquired in advance, and the resin type that does not contain carbon black and whose spectral data is equal to or greater than a second threshold set by a preset, first similarity is determined to be the resin type of the object to be sorted, A resin determination method in which, if the reflection or absorption spectrum obtained by the reflected light is less than the first threshold, a second similarity is determined with spectral data of two or more resin types, including carbon black-containing resin types and carbon black-free resin types, that have been acquired in advance, and the carbon black-containing resin type or carbon black-free resin type whose second similarity is equal to or greater than a predetermined third threshold of second similarity and whose spectral data is the highest, is determined to be the resin type of the object to be sorted.

2. The resin determination method according to claim 1, wherein the effective wavelength range of the infrared light is 1 μm or more and 3 μm or less.

Images