Apparatus, sorting machine, and method for automatic sorting machines

Abstract

This specification discloses an apparatus (10) for an automatic sorting machine (100), the apparatus enabling the automatic sorting machine to sort a substance (12) to a target specification, the apparatus comprising a sensor device (14) adapted to take readings of the substance (12) passing through a sensor device, and at least one processing device (18) which receives a first classifier (T) that defines an acceptable range of properties of the substance to satisfy the target specification, and optionally receives at least one second classifier (M1, M2) that defines an acceptable range, detects the properties of the substance based on the taken readings, and determines whether the detected properties are within the acceptable range of the first classifier or at least one second classifier The system comprises at least one processing device (18) configured to determine whether the detected characteristics are within an acceptable range or outside the range of the first and second classifiers, to accept the substance (12') if the detected characteristics are within an acceptable range, to accept the substance (12'') if the current average value of the characteristics of the sorted substances, updated with the detected characteristics, is within an acceptable range if the detected characteristics are within an acceptable range, to reject the substance (12''') if the current average value of the characteristics, updated with the detected characteristics, is outside an acceptable range, and to reject the substance (12'''') if the detected characteristics are outside the range of the first and second classifiers.

Description

[Technical Field]

[0001] This disclosure relates to equipment for an automated sorting machine. This disclosure also relates to an automated sorting machine equipped with such equipment. This disclosure also relates to a method for sorting materials to a target specification. [Background technology]

[0002] Automatic sorting machines can be used for sorting waste for recycling purposes. For example, automatic sorting machines can be used to sort a stream of plastic or metal waste, such as plastic or metal flakes or chips.

[0003] An exemplary automated sorter for flakes may comprise a hopper for receiving the material / flakes to be sorted, a scanner box for material and color detection, an electromagnetic sensor for metal detection, an air discharge means with valves and nozzles, and a separation chamber. Clear / light blue (PET) flakes are accepted while impurities (colored flakes, non-PET flakes, and metals) are discharged. Rejected flakes may be scanned a second time.

[0004] European Patent No. 2832458 provides an optical granular sorter that enables sensitivity setting by utilizing RGB three-dimensional color space information similar to information obtained through the human eye. Data is created for the wavelength components of R, G, and B light in the three-dimensional color space from the granular material. The creation unit sets an interface calculated based on the Mahalanobis distance to divide the data into suitable granular material cluster regions and unsuitable granular material cluster regions. The Euclidean distance interface creation unit determines the centroid positions of the suitable granular material cluster regions and the unsuitable granular material cluster regions, and sets an interface calculated based on the Euclidean distance where the centroid positions are furthest apart. The threshold determination unit determines the intersection line between the interface calculated based on the Mahalanobis distance and the interface calculated based on the Euclidean distance, and determines that the intersection line becomes a discrimination threshold that can determine whether or not to treat the granular material as a target for separation. [Overview of the Initiative]

[0005] Therefore, an object of this disclosure is to provide an improved apparatus that enables an automated sorting machine to sort materials to target specifications, such as a target color or target recipe for an alloy.

[0006] To achieve the above-mentioned objectives, and other objectives that may become apparent from the following description, the present disclosure provides an apparatus having the features defined in claim 1. Preferred modifications of the apparatus will become apparent from the dependent claims.

[0007] More specifically, according to a first aspect, an apparatus for an automatic sorting machine is provided, the apparatus enabling the automatic sorting machine to sort materials to a target specification, the apparatus comprising a sensor device adapted to take readings of materials passing through the sensor device, and at least one processing device, a) Receiving a first input of a first classifier that defines an acceptable range of material properties to satisfy the target specifications, b) To satisfy the target specification, receive a second input of at least one second classifier which defines an potentially acceptable range of the characteristics, c) Based on the acquired readings, the properties of the substance are detected, d) Determine whether the detected characteristics are within the acceptable range of the first classifier, or within the potentially acceptable range of at least one second classifier, or outside the range of both the first and second classifiers. e) If the detected characteristics are within the acceptable range of the first classifier, the material is accepted by the automatic sorting machine. f) If the detected characteristics are within the potentially acceptable range of at least one second classifier, and the current average value of the characteristics of the sorted material, updated with the detected characteristics, is within the acceptable range of the first classifier, the material is allowed to be accepted by the automatic sorter; if the current average value of the characteristics, updated with the detected characteristics, is outside the acceptable range of the first classifier, the material is rejected by the automatic sorter. g) If the detected characteristics fall outside the range of the first and second classifiers, at least one processing apparatus configured to cause the substance to be rejected by an automatic sorter, It is equipped with.

[0008] This disclosure is based on the understanding that by sorting substances based on mixing rules such as those in features e) to g), that is, by accepting substances within an acceptable range of a first classifier, but also accepting substances within an potentially acceptable range of at least one second classifier if the average properties of the sorted substances become acceptable (or otherwise the substances are rejected), fewer substances may be rejected and better mixing efficiency may be achieved.

[0009] The sensor device may be adapted to capture readings of a substance passing through the sensor device at a speed of preferably 0.4 m / s to 20 m / s, either on a conveyor belt, on a chute, or in free fall.

[0010] The sensor device may be adapted to continuously acquire readings of batches of material passing through the sensor device, and at least one processing device is configured to perform steps c) to g) (continuously / repeatedly) for substantially each material in the batch. In this way, bulk sensor-based sorting to target specifications can be achieved.

[0011] At least one processing unit may be configured to update the current average value of the properties of the sorted material with the detected properties of the accepted material. Specifically, at least one processing unit may be configured to update the current average value of the properties of the sorted material in a batch using vectors and functions.

number

[0012] where v’ b is the updated current average value of the characteristic, v b is the current average value of the characteristic, v is the detected characteristic of the received substance, and r is a ratio within the range of 10 -308 to 0.5. That is, here, the sorting is based on the averaging vector. If r is very small, more inferior-quality materials may be accepted into the batch. If r is higher, the sorting becomes more stringent. Here, since the batch operation is performed on the vector, no special logic for the classifier is required. Furthermore, the execution time can be shortened.

[0013] In one embodiment, the target specification includes a target color, and the detected characteristic includes the color of the substance. The target color can be, for example, transparent / light blue.

[0014] At least one processing device may be configured here to represent the detected color as coordinates within a color circle color space, and the first and second classifiers are polygons within the color circle color space. Specifically, the detected color can be described using the hue (H) and saturation (S) represented by the color circle coordinates. The advantages of using the circular representation of HS include that all gray tones (from black to white) are in the center, there are no singularities for dark and unsaturated colors, there are no problems with the visualization of hue 0° and 359°, the mathematics is easier to mix, there are no problems at 0° / 359°, and the circular coordinates can be quickly converted by the dot product. Color spaces other than the HS(V) color circle, such as RGB, IHS, CIE L*a*b*, CIE L*u*v*, YUV, YIQ, HSV, xy, rg, HSI, HLS, YCbCr, OHTA, LCH-uv, LCH-ab, etc., can also be used in the same way.

[0015] In another embodiment, the target specification includes a target alloy composition (or "recipe"), and the detected characteristic includes the elemental composition of the substance.

[0016] The application of the disclosed method to aluminum alloys (and their recycling) can, advantageously, enable less waste generated during batches, and therefore a higher recycling rate.

[0017] Furthermore, it can provide a greater level of flexibility to the alloy recipes applied to the recycling process.

[0018] The sensor device may include a digital RGB (red, green, blue) camera, which is adapted to take an image of a substance passing through the digital RGB camera (i.e., "capture readings" in claim 1), and the detected characteristic is the color of at least one pixel of the substance in the image. At least one processing device may be configured, hereby using a dot product, preferably an average normalized dot product, to convert the RGB values ​​of the at least one pixel to the coordinates in the color circle color space. This conversion from RGB values ​​to (color circle) coordinates is rapid and contributes to the rapid sorting of the substance.

[0019] The sensor device may optionally or additionally include a visual spectrometer adapted to acquire visual spectral data of a substance (= "acquire readings"), and at least one processing unit configured to detect the color of the substance based on the acquired visual spectral data. The at least one processing unit can here be configured to first convert the visual spectral data to RGB using CIE, and then convert the RGB to (color circle) coordinates. Furthermore, the spectrometer output can be used as a vector for averaging without conversion.

[0020] The sensor device may also be configured to perform laser-induced breakdown spectroscopy (LIBS) as an alternative or additional measure, which enables accurate determination and quantification of alloying elements and thus allows for differentiation between a number of different alloy types.

[0021] At least one processing device may be further configured to detect the intensity (g) of a substance based on the acquired readings and allow the automatic sorter to accept the substance if the detected characteristic (e.g., color) is within an acceptable range of a first classifier and the detected intensity is within a first predetermined acceptable intensity range. At least one processing device may be further configured to allow the automatic sorter to accept the substance if i) the detected characteristic is within an optionally acceptable range of at least one second classifier, ii) the current average value of the characteristics (of the sorted substance) updated with the detected characteristic is within an acceptable range of the first classifier, iii) the detected intensity is within a second predetermined optionally acceptable intensity range, and iv) the current average value of the intensity of the sorted substance updated with the detected intensity is within a first predetermined acceptable intensity range. By also taking into account the intensity or brightness of the substance, more precise sorting can be achieved. The intensity of a substance can be detected / determined from the readings (image / visual spectral data) using dot product calculations.

[0022] The first predetermined acceptable intensity range is preferably a narrower subrange of the second predetermined, optionally acceptable intensity range. In this way, the apparatus can tolerate more deviation, but the mixture must not exceed the (narrower) first predetermined acceptable intensity range.

[0023] In another embodiment, the target specification includes a target material type, the sensor device comprises a spectrometer, the spectrometer is adapted to determine (= "capture readings") the spectrum of a substance passing through the spectrometer, and the detected characteristic is the material type of the substance. The target material type may be, for example, polyethylene terephthalate (PET), PE, PO, PVC, etc. Here, at least one processing device may be configured to represent the detected material (type) as coordinates in a scatter plot, and the first and second classifiers may be polygons in the scatter plot.

[0024] In further embodiments, the target specification may include both a target color and a target material type, where a first classifier defines an acceptable range of color for a substance to satisfy the target specification, and another first classifier defines an acceptable range of material type for a substance to satisfy the target specification, and at least one processing apparatus is configured to allow the automatic sorter to accept the substance if the detected color is within the acceptable range of the first classifier and the detected material type is within the acceptable range of the other first classifier. Furthermore, at least one processing apparatus may be configured to allow the automatic sorter to accept the substance if i') the detected color is within an optionally acceptable (color) range of at least one second classifier, ii') the current average value of the color updated with the detected color is within an acceptable range of the first classifier, iii') the detected material type is within a second predetermined optionally acceptable material type range, and iv') the current average value of the material type of the sorted substance updated with the detected material type is within an acceptable range of the other first classifier. Furthermore, at least one processing apparatus may be configured to update the current average value of the color of the sorted material by the detected color of the accepted material, and the current average value of the material type of the sorted material by the detected material type of the accepted material (both). Furthermore, a material may be accepted if one of the detected color and material type is in its first classifier and the other is in its second classifier, and the updated current average value of the other is in its first classifier. In one embodiment, if one of the detected color and material type is in its second classifier and the updated current average value is in its first classifier, the other must be in its first classifier for a material to be accepted (i.e., in this embodiment, materials satisfying i')~iv') are not accepted).Furthermore, at least one processing apparatus can be configured to cause an automated sorter to reject a substance if one (or both) of the detected color and material type are outside the respective first classifiers, or if all criteria i') to iv') are not met. Here, the sensor apparatus may comprise the aforementioned spectrometer and (both) a digital RGB camera or a visual spectrometer.

[0025] Generally, the target specification may include a first characteristic (e.g., color) and a second different characteristic (e.g., material type), where a first classifier defines an acceptable range of the first characteristic of a substance to satisfy the target specification, and another first classifier defines an acceptable range of the second characteristic of a substance to satisfy the target specification, and at least one processing device is configured to detect the first and second characteristics of the substance based on the acquired readings and optionally / possibly based on at least one other reading of the substance acquired by a sensor device (e.g., a digital RGB camera and a spectrometer), and to allow the substance to be accepted by an automatic sorter if the detected first characteristic is within the acceptable range of the first classifier and the detected second characteristic is within the acceptable range of the other first classifier.

[0026] According to a second aspect of the present disclosure, an automated sorting machine is provided which is adapted to sort materials to a target specification, the automated sorting machine comprising an apparatus according to the first aspect, means for providing a flow of materials through a sensor device of the apparatus, a container adapted to receive the received materials from the flow, and a discharge device adapted to remove rejected materials from the flow before they reach the container. The means for providing a flow of materials through the sensor device comprises at least one of a conveyor belt, a chute, and a free-fall device. The means may be configured to operate at a speed of 0.4 m / s to 20 m / s.

[0027] A third aspect of the present disclosure provides a method for classifying a substance to a target specification, the method comprising: receiving a first input (e.g., from a user interface) of a first classifier defining an acceptable range of properties of the substance to satisfy the target specification; receiving a second input (e.g., from a user interface) of at least one second classifier defining an potentially acceptable range of properties to satisfy the target specification; capturing readings of the substance passing through a sensor device by a sensor device; detecting the properties of the substance based on the captured readings; and determining by at least one processing device whether the detected properties are within the acceptable range of the first classifier or in the case of at least one second classifier. Therefore, the method includes the steps of determining whether the detected characteristic is within an acceptable range or outside the range of the first and second classifiers; if the detected characteristic is within the acceptable range of the first classifier, allowing the material to be accepted by an automatic sorter; if the detected characteristic is within an acceptable range of at least one second classifier, allowing the material to be accepted by an automatic sorter if the current average value of the characteristics of the sorted material updated with the detected characteristic is within the acceptable range of the first classifier, and allowing the material to be rejected by the automatic sorter if the current average value of the characteristics updated with the detected characteristic is outside the acceptable range of the first classifier; and if the detected characteristic is outside the range of the first and second classifiers, allowing the material to be rejected by the automatic sorter. This embodiment may exhibit the same or similar features and technical effects as the first and / or second embodiment, and vice versa.

[0028] The substance may be at least one plastic flake, such as PET flakes, PE flakes, PO flakes, and / or PVC flakes, or may contain them.

[0029] Furthermore, the step of having the automatic sorter accept the material may include refraining from sending a discharge signal to the discharge device of the automatic sorter for the material, thereby receiving the material into a first container of the automatic sorter. Furthermore, the step of having the automatic sorter reject the material may include sending a discharge signal to the discharge device for the material. Here, the method may further include the step of the discharge device discharging the material based on the discharge signal, thereby receiving the material into a second container of the automatic sorter.

[0030] One or more embodiments of this disclosure will be described, by reference to the following figures, merely as examples. [Brief explanation of the drawing]

[0031] [Figure 1] A schematic side view of an apparatus according to one aspect of this disclosure. [Figure 2] This is a schematic side view of an automatic sorting machine according to another aspect of the present disclosure. [Figure 3] This is a flowchart of a method according to yet another aspect of the present disclosure. [Figure 4] This shows the classifiers in the color circle color space. [Figure 5] An example GUI for entering classifiers is shown. [Figure 6] This is a flowchart of a modified example of the method shown in Figure 3. [Figure 7] It indicates a specified intensity range. [Figure 8] The material type classifiers in the scatter plot are shown. [Modes for carrying out the invention]

[0032] This disclosure is illustrated below by several illustrative examples. It will be understood that these examples are provided for illustrative and illustrative purposes only and are not intended to limit the scope of this disclosure. Instead, the scope of this disclosure is defined by the appended claims.

[0033] Furthermore, while embodiments are presented individually for focused discussion of specific features, it will be recognized that this disclosure also encompasses combinations of embodiments described herein.

[0034] Figure 1 shows a device 10 for an automatic sorting machine 100 (see Figure 2). Thus, the device 10 may be included in the automatic sorting machine 100. The device 10 is generally adapted to enable the automatic sorting machine 100 to sort material 12 to a target specification, for example, a target color such as clear / light blue. That is, the device 10 should enable the automatic sorting machine 100 to accept / collect only material 12' that contributes to meeting the target specification, while material 12''' that similarly enters the machine 100 but does not contribute to the target specification is rejected / discarded. Material 12 may include, for example, plastic flakes of different colors from the recycling of packaging / containers, typically such as PET bottles.

[0035] The apparatus 10 includes a sensor device 14. The sensor device 14 is adapted to capture readings of the substance 12 passing through the sensor device 14. The sensor device 14 may include, for example, a digital RGB camera adapted to capture a digital image (e.g., a video frame) of the substance 12 passing through a digital RGB camera, or a visual spectrometer adapted to capture visual spectral data of the substance 12 passing through the sensor device 14.

[0036] The substance 12 passing through the sensor device 14 can form a flow of substance 12 as indicated by the arrow 16. For this purpose, the digital RGB camera or visual spectrometer of the sensor device 14 may be directed toward the flow of substance 12. The flow of substance 12 passing through the sensor device 14 may be provided, for example, by a conveyor belt or chute 102 of an automatic sorting machine 100 (see Figure 2). During operation, the substance 12 (flow) can pass through the sensor device 14 at a speed of, for example, 0.4 m / s to 20 m / s.

[0037] The apparatus 10 further comprises at least one processing unit 18. The at least one processing unit 18 may be connected to the digital RGB camera or vision spectrometer described above. The at least one processing unit 18 may include, for example, a circuit / processor (CPU), in particular the processor (CPU) of the computer (of the apparatus 10). The at least one processing unit 18 is configured (e.g., by software) to perform various tasks or steps, as described below with further reference to Figure 3. Figure 3 shows a possible way of operating the apparatus 10.

[0038] In a), at least one processing unit 18 receives a first input of the first classifier T. The first input of the first classifier T can be received from a user interface, for example, a graphical user interface (GUI) 200 shown in Figure 5. The user interface 200 is typically located away from the device 10 and connected to the device 10 via a wired or wireless connection 202. Furthermore, the first classifier T is typically input by a human operator via the user interface 200.

[0039] The first classifier T defines the acceptable range of properties of material 12 to satisfy the target specification. If the target specification is a target color, the first classifier T can define an acceptable color range. As will be further explained below, the first classifier T may be, for example, a closed shape within the color circle color space 20, i.e., a polygon (see Figures 4-5). If the target color is transparent / light blue, the first classifier T may be a polygon at the center of the color circle color space 20, as shown in Figures 4-5. In Figure 4, different patterns schematically represent different colors in the color circle color space 20.

[0040] In (b), at least one processing unit 18 receives a second input of at least one second classifier M1, M2. The first second input of at least one second classifier M1, M2 can be received from the user interface 200. The at least one second classifier M1, M2 is typically input by a human operator via the user interface 200.

[0041] Alternatively, classifiers such as T, M1, and M2 can be automatically configured by recording the training set and using supervised learning to automatically fit the classifiers.

[0042] At least one second classifier M1, M2 defines an optionally acceptable range of properties of substance 12 in order to satisfy the target specification. If the target specification is a target color, at least one second classifier M1, M2 may define an optionally acceptable color range, which may also be called an acceptable mixed candidate color. At least one second classifier M1, M2 may be, for example, one or more closed figures in the color circle color space 20, i.e., at least one polygon. If the target color is transparent / light blue, the second classifier M1 may be, for example, a polygon enclosing the (smaller) first classifier T at the center of the color circle color space 20, as shown in Figures 4-5, and the second classifier M2 may be a polygon located in the blue region of the color circle color space 20. Therefore, the first classifier T may be called a strict classifier, and at least one second classifier M1, M2 may be called at least one relaxed classifier.

[0043] Please note that b) can be done before, after, or simultaneously with a).

[0044] In c), at least one processing device 18 detects the relevant characteristic (e.g., color) of individual substances 12 (e.g., plastic flakes) based on readings acquired by the sensor device 14. If the target specification is a target color, the detected color of the substance 12 may be expressed as (color circle) coordinates in the color circle color space 20. Specifically, the detected color can be described using hue (H) and saturation (S) represented by color circle coordinates.

[0045] If the sensor device 14 includes a digital RGB camera, the detected characteristic may be the color of at least one pixel of the substance 12 in the image captured by the digital RGB camera, and at least one processing device 18 can convert the RGB value of at least one pixel into the corresponding (color circle) coordinates in the color circle color space 20 using a dot product operation (color circle), preferably an average normalized dot product. Expressed in vector notation, this is the conversion from RGB values ​​to color circle coordinates (c x ,c y The conversion to ) may be as follows:

number

[0046] When R, G, and B are mean-normalized, scaling is unnecessary. Therefore, for mean-normalized RGB values, the following transformation can be used.

number

[0047] When the sensor device 14 includes a visual spectrometer, at least one processing device 18 can detect the color of the substance 12 based on the captured visual spectral data. The at least one processing device 18 can be configured here to first convert the visual spectral data to RGB using the CIE (step 1), and then convert the RGB to color circle coordinates (step 2).

[0048] In step 1, for example, the CIE matching functions (fr, fg, fb) from the table on p. 750 of "Color Science" (1982) by Wyszecki and Stiles can be used to calculate the typical RGB stimuli for a human observer. If the visual spectrometer is clipped in the blue channel, a white balance can be used on the natural image; otherwise, the blue may be underrepresented. The score vector can then be scaled to the range 0 - 255.

[0049] In step 2, the above - mentioned conversion from RGB to color circle coordinates uses replacing R, G, B with f r , f g , f b . The vector data must be mean - normalized.

Equation

[0050] where v μ T is the mean - normalized spectrum.

[0051] Matrix multiplication results in the following calculation using two score vectors S x , S y .

Equation

[0052] In d), at least one processing device 18 determines that the detected characteristic is - Within the acceptable range of the first classifier T (e.g., the detected color c of the substance) x , c y (However, it is inside the first classifier / polygon T), - Within an acceptable range in some cases for at least one second classifier (e.g., the detected color c of the substance) x , c y (However, it is located inside either the second classifier / polygon M1 or M2), or - Outside the range of the first and second classifiers (e.g., the detected color of the substance c) x , c y However, are they located outside the first and second classifiers / polygons T, M1 and M2? Determine.

[0053] If the detected characteristics are within the acceptable range of the first classifier, for example (c x , c y If ) ∈ T, then at least one processing unit 18 may cause the automatic sorter 100 to accept the individual substance designated hereby 12' in e). To do this, at least one processing unit 18 may refrain from sending a discharge signal to the discharge device 104 of the automatic sorter 100 for this substance 12', and as a result, this substance 12' is not removed from the flow 16 but is instead received into the container 106 of the automatic sorter 100.

[0054] If the detected characteristics are within an acceptable range for at least one second classifier, for example (c x , c y ) ∈ M1 or (c x , c y If ) ∈ M2, at least one processing unit 18 also updates the batch B of substances with the detected properties of the individual substances designated here as 12'', and the current average value of the properties of the already sorted (accepted) substances is also the first classifier T (e.g., (c x ', c yEven if ') ∈ T) is within the acceptable range, the specified individual substances 12'' can be accepted by the automatic sorting machine 100 at f), where the update may be as follows:

number

[0055] In other words, vector c x ,c y This must match the colors already seen in batch B. Therefore, the acceptable range of the first classifier T can also be called the acceptable average color.

[0056] On the other hand, if the current average value of the traits updated with the detected traits is outside the acceptable range of the first classifier T, for example,

number

[0057] If the detected characteristics fall outside the range of the first and second classifiers T, M1 and M2, at least one processing unit 18 can cause the automatic sorter 100 to reject the individual substance designated hereby 12''''. To do this, at least one processing unit 18 can, in h2), send a discharge signal for this substance 12'''' to the discharge unit 104, so that the substance 12'''' is removed from the flow 16 and not received in the container 106.

[0058] By selecting substances based on mixing rules as described in features e) to g), that is, by accepting substances within the acceptable range of the first classifier T, but also accepting substances within the potentially acceptable range of at least one second classifier M1, M2 if the average properties of the selected substances become acceptable (or otherwise the substances are rejected), fewer substances may be rejected, and better mixing efficiency may be achieved.

[0059] When substance 12' or 12'' is accepted in e) or f), at least one processing unit 18 can update the current average value of the properties of the sorted substances in batch B with the detected properties of the accepted substance 12' or 12'' in each of i1) or i2). This update is generally performed

number

number

[0060] An alternative update function that imposes a higher penalty depending on the vector distance to the center is as follows:

number

[0061] At least one processing unit 18 may be configured to perform c) to g) (and h1), h2), i1), i2)) for substantially each of the aforementioned batches, if applicable, as shown by arrow 22 in Figure 3 (parallel processing is assumed). Furthermore, the initial starting value of the current average of the properties of the sorted materials for the batch may be a predetermined starting value (in the case of a vector). Furthermore, the current average of the properties of the sorted materials may be reset to a predetermined starting value, for example, at startup, after a break, or at the user's request.

[0062] As described above, the automatic sorting machine 100 in Figure 2 comprises an apparatus 10 including a sensor device 14 and at least one processing device 18; a conveyor belt or chute 102 for providing a flow 16 of material 12 that passes through the sensor device 14 toward a container 106; a container 106 adapted to receive the accepted material 12', 12'' from the flow 16; and a discharge device 104 adapted to remove rejected material 12''', 12'''' from the flow 16 before reaching the container 106.

[0063] The conveyor belt or chute 102 may be configured to operate at a speed of 0.4 m / s to 20 m / s.

[0064] The discharge device 104 may be connected to at least one processing device 18 / device 10 to receive the aforementioned discharge signal. The discharge device 104 may be equipped with a plurality of air nozzles 108 distributed along the width of the conveyor belt or chute 102 to blow (air discharge) the rejected material 12'', 12'''' away from the flow 16, preferably into an additional container 110 of the automatic sorter 100. The containers 106, 110 may be called the first and second containers and / or separation chambers.

[0065] The automatic sorting machine 100 may also include at least one hopper 112 for feeding the material 12 onto a conveyor belt or chute 102. The automatic sorting machine 100 may also include an electromagnetic sensor 114 for metal detection, allowing any metal in the flow 16 to be removed by the discharge device 104.

[0066] Looking at Figures 6 and 7, the method in Figure 6 is similar to the method in Figure 3, but in c'), it further includes at least one processing device 18 of the apparatus 10 that detects the intensity g of individual substances 12 based on the acquired readings.

[0067] The intensity of a substance can be detected / determined from readings (images from an RGB camera / visual spectral data from a spectrometer) using dot product calculations. Intensity can be defined as follows:

number

number

[0068] In the case of a visual spectrometer (16 channels), intensity can be defined as follows:

number

[0069] In e'), at least one processing apparatus 18 determines that the detected properties of the substance (e.g., color) are within the acceptable range of the first classifier T (as shown in Figure 3), and in c'), the detected intensity is within the first predetermined acceptable intensity range t1~t2, for example,

number

[0070] In f'), at least one processing device 18 can cause the automatic sorter 100 to accept individual substances 12'' if i) the detected characteristic (e.g., color) is within an acceptable range of at least one second classifier T (as shown in Figure 3), ii) the current average value of the characteristics updated with the detected characteristic is within an acceptable range of the first classifier T (as shown in Figure 3), iii) the detected intensity is within a second predetermined and possibly acceptable intensity range m1~m2 (m1≦g≦m2), and iv) the current average value of the intensity of the already sorted (accepted) substances in batch B of substance 12, updated with the detected intensity of the individual substances 12', is within a first predetermined acceptable intensity range t1~t2 (t1≦g'≦t2). Here, the updates may be as follows:

number

[0071] A second, in a predetermined case, acceptable intensity range m1 to m2 can be set, for example, by the operator using the user interface 200. The second, in a predetermined case, acceptable intensity range m1 to m2 may be, for example, 50 to 250*. Thus, as can be seen in Figure 7, the first, in a predetermined case, acceptable intensity range t1 to t2 may be a narrower subrange of the second, in a predetermined case, acceptable intensity range m1 to m2.

[0072] If criteria i) to iv) are not all met, at least one processing device 18 may cause the automatic sorting machine 100 to reject the substance, for example, by transmitting an discharge signal in h1).

[0073] When substance 12' or 12'' is accepted in e') or f'), at least one processing device 18 can also update the current average intensity of the sorted substances in batch B with the detected intensity of substance 12' or 12'' accepted in i1') or i2'), respectively.

number

[0074] A substance may be accepted if one of the detected properties (e.g., color) and / or detected intensity is in its first classifier, the other is in its second classifier, and the updated current average value of the other is in its first classifier. Furthermore, in one embodiment, if one of the detected properties and / or detected intensity is in its second classifier, and the updated current average value is in its first classifier, the other must be in its first classifier for an accepted substance (i.e., in this embodiment, substances satisfying i) to iv) are not accepted).

[0075] By also taking into account the intensity or brightness of substance 12, more precise sorting can be achieved.

[0076] Referring to Figure 8, in another embodiment, instead of color, the target specification may be a target material (type), and the sensor device 14 comprises a spectrometer adapted to determine the spectrum of substance 12 passing through the spectrometer, and the detected characteristic is the material type of the substance. The target material type may be, for example, PET, PE, PO, PVC, etc. Here, at least one processing device 18 may be configured to represent the detected material type as coordinates (i.e., points) in a scatter plot 24, and the first and second classifiers T', M' are polygons in the scatter plot 24. Specifically, spectral data containing the spectrum of substance 12 may be normalized using a standard normal variate (SNV), and principal component analysis (PCA) may be performed to project the spectral data onto the scatter plot 24 using the first and second principal axes. In the specific example shown in Figure 8, T' is the first classifier for sorting PE-HD (high-density polyethylene), and M' is the second classifier for potentially accepting PE-LD (low-density polyethylene) as well.

[0077] In another example, the flow of a substance can be analyzed, for example using LIBS, to determine its composition, which can then be compared to a target composition or "recipe" for an alloy (e.g., an aluminum alloy).

[0078] The second classifier in this example can be defined by a fixed maximum acceptable contamination level for each element and / or an adaptive acceptable contamination level for each element, taking into account the batch's input stream composition and maturity over time.

[0079] For example, in the initial stage, only a fixed maximum value may be used, and within the batch, a moderate level of contamination may be tolerated. At the end of the batch, everything can be filled up to the recipe's limit. Depending on the implementation, a similar strategy can be used to maximize the mass in the recipe from the input stream.

[0080] Because LIBS measurements are performed over a very small area, different strategies can be applied to represent batches. Batches and particles can be represented by a single measurement point (only at the measurement location), a measurement point multiplied by the area of ​​the particle, and / or a measurement point multiplied by the weight (volume + density) of the particle, with heuristic correction factors optionally applied to the measurements.

[0081] Adding particles to a batch may involve spectrally accumulating the batch to average the LIBS spectrum and accumulating elemental composition vectors (e.g., including copper, magnesium, silicon, iron, lead, zinc, etc.).

[0082] While various modifications and alternative forms are possible with respect to this disclosure, specific embodiments are shown and described as examples in relation to the drawings in order to clearly illustrate the various advantageous aspects of this disclosure. However, it should be understood that the detailed description herein and the drawings accompanying this disclosure are not intended to limit this disclosure to any particular form disclosed. Rather, they are intended to cover all modifications, equivalents, and alternatives that fall within the following claims.

[0083] For example, other color spaces besides the HS(V) color circle, such as RGB, IHS, CIE L*a*b*, CIE L*u*v*, YUV, YIQ, HSV, xy, rg, HSI, HLS, YCbCr, OHTA, LCH-uv, and LCH-ab, can be used in the same way.

[0084] Furthermore, multiple sensors and spectrometers can be combined by checking all requirements from all sensors before accepting and updating pixels (for example, sorting them to a target color and target material).

Claims

1. An apparatus (10) for an automatic sorting machine (100), wherein the apparatus enables the automatic sorting machine to sort substances (12) to a target specification, and the apparatus A sensor device (14) adapted to capture readings of substances (12, 12', 12'', 12'''', 12'''') passing through the sensor device, At least one processing device (18), a) Receiving a first input of a first classifier (T) that defines an acceptable range of properties of the material in order to satisfy the target specifications, b) At least one second classifier (M) that defines an acceptable range of the characteristics in some cases in order to satisfy the target specifications. １ M ２ ) receives the second input, c) Based on the acquired readings, the characteristics of the substance are detected, d) Determine whether the detected characteristic is within the acceptable range of the first classifier, within the potentially acceptable range of the at least one second classifier, or outside the range of the first and second classifiers. e) If the detected characteristics are within the acceptable range of the first classifier, the automatic sorting machine will accept the substance (12'), f) If the detected characteristic is within the potentially acceptable range of the at least one second classifier, and the current average value of the characteristics of the sorted material, updated with the detected characteristic, is within the acceptable range of the first classifier, the automatic sorter will accept the material (12''); if the current average value of the characteristics, updated with the detected characteristic, is outside the acceptable range of the first classifier, the automatic sorter will reject the material (12'''). g) If the detected characteristics fall outside the range of the first and second classifiers, the automatic sorting machine rejects the substance (12''''). A device comprising at least one processing unit (18) and A device (10) comprising:

2. The apparatus according to claim 1, wherein the sensor device is adapted to capture readings of a substance passing through the sensor device at a speed of preferably 0.4 m / s to 20 m / s, on a conveyor belt, on a chute (102), or in free fall.

3. The apparatus according to claim 1 or 2, wherein the sensor device is adapted to take readings of a batch of substance passing through the sensor device, and the at least one processing device is configured to perform c) to g) substantially for each substance in the batch.

4. The apparatus according to any one of claims 1 to 3, wherein the at least one processing apparatus is configured to update the current average value of the properties of the sorted material with the detected properties of the accepted material.

5. The aforementioned at least one processing device processes vectors and functions, i.e., [Math 1] It is configured to update the current average value of the properties of the sorted material in the batch using the following: In the formula, v' ｂ v is the updated current average value of the aforementioned characteristics, ｂ v is the current average value of the aforementioned characteristics, v is the detected characteristic of the accepted substance, and r is 10 －３０８ The apparatus according to claims 3 and 4, wherein the ratio is within the range of ~0.

5.

6. The apparatus according to any one of claims 1 to 5, wherein the target specification includes a target mixture of elements, and the detected properties include the elemental composition of the substance.

7. The apparatus according to claim 6, wherein the second classifier includes a fixed maximum acceptable contamination level or an adaptive acceptable contamination level for each element.

8. The apparatus according to any one of claims 1 to 5, wherein the target specification includes a target color, and the detected characteristic includes the color of the substance.

9. The apparatus according to claim 8, wherein the at least one processing device is configured to represent the detected color as coordinates in a color circle color space (20), and the first and second classifiers are polygons in the color circle color space.

10. The apparatus according to claim 8 or 9, wherein the sensor device comprises a digital RGB camera, the digital RGB camera is adapted to take an image of the substance passing through the digital RGB camera, and the detected characteristic is the color of at least one pixel of the substance in the image.

11. The apparatus according to claim 10, wherein the at least one processing unit is configured to use a dot product operation to convert the RGB values ​​of the at least one pixel to the coordinates in the color circle color space.

12. The apparatus according to any one of claims 8 to 11, wherein the sensor device comprises a visual spectrometer adapted to acquire visual spectral data of the substance, and the at least one processing device is configured to detect the color of the substance based on the acquired visual spectral data.

13. The apparatus according to any one of claims 1 to 12, wherein the at least one processing device is configured to detect the intensity of the substance based on the acquired readings, and if the detected characteristic is within the acceptable range of the first classifier and the detected intensity is within a first predetermined acceptable intensity range (t1 to t2), the automatic sorter accepts the substance; i) if the detected characteristic is within the optionally acceptable range of the at least one second classifier, ii) if the current average value of the characteristics updated with the detected characteristic is within the acceptable range of the first classifier, iii) if the detected intensity is within a second predetermined optionally acceptable intensity range (m1 to m2), and iv) if the current average value of the intensity of the sorted substance updated with the detected intensity is within a first predetermined acceptable intensity range (t1 to t2).

14. The apparatus according to claim 13, wherein the first predetermined acceptable strength range (t1 to t2) is a narrower subrange of the second predetermined, in some cases acceptable strength range (m1 to m2).

15. The apparatus according to any one of claims 1 to 14, wherein the target specification includes a target material type, the sensor device comprises a spectrometer, the spectrometer is adapted to determine the spectrum of the substance passing through the spectrometer, and the detected characteristic is the material type of the substance.

16. The apparatus according to claims 8 and 15, wherein the target specification includes a target color and a target material type, the first classifier (T) defines an acceptable range of color for the substance to satisfy the target specification, another first classifier (T') defines an acceptable range of material type for the substance to satisfy the target specification, and the at least one processing apparatus is configured to cause the automatic sorter to accept the substance if the detected color is within the acceptable range of the first classifier and the detected material type is within the acceptable range of the other first classifier.

17. Apparatus according to any one of claims 1 to 16, wherein the target specification includes a first characteristic and a second distinct characteristic, the first classifier (T) defines an acceptable range of the first characteristic of the substance to satisfy the target specification, and another first classifier (T') defines an acceptable range of the second characteristic of the substance to satisfy the target specification, and the at least one processing device is configured to detect the first and second characteristics of the substance based on the acquired reading and optionally based on at least one other reading of the substance acquired by the sensor device, and to cause the automatic sorter to accept the substance if the detected first characteristic is within the acceptable range of the first classifier and the detected second characteristic is within the acceptable range of the other first classifier.

18. An automatic sorting machine (100) adapted to sort substances (12) to target specifications, The apparatus (100) according to any one of claims 1 to 17, Means (102) for providing a flow of substance (16) passing through the sensor device (14) of the apparatus, A container (106) adapted to receive the received substance from the flow, Discharge device (104) adapted to remove rejected material from the flow before it reaches the container (106) and An automatic sorting machine (100) equipped with the following.

19. The automatic sorting machine according to claim 18, wherein the means for providing a flow of material through the sensor device comprises at least one of a conveyor belt, a chute, and a free-fall device.

20. A method for sorting a substance (12) to a target specification, The steps include receiving a first input of a first classifier (T) that defines an acceptable range of properties for the material in order to satisfy the target specifications, To satisfy the aforementioned target specifications, at least one second classifier (M) defines an acceptable range for the characteristics in some cases. １ M ２ The step of receiving a second input of ) The sensor device (14) acquires the reading of the substance passing through the sensor device, A step of detecting the properties of the substance based on the acquired readings, A step of determining by at least one processing device (18) whether the detected characteristic is within the acceptable range of the first classifier, within the optionally acceptable range of the at least one second classifier, or outside the ranges of the first and second classifiers, If the detected characteristics are within the acceptable range of the first classifier, the step is to allow the material to be accepted by an automatic sorting machine. If the detected characteristic is within the potentially acceptable range of the at least one second classifier, and the current average value of the characteristics of the sorted material updated with the detected characteristic is within the acceptable range of the first classifier, the automatic sorter is allowed to accept the material; if the current average value of the characteristics updated with the detected characteristic is outside the acceptable range of the first classifier, the automatic sorter is allowed to reject the material. If the detected characteristics fall outside the range of the first and second classifiers, the automatic sorting machine is used to reject the substance. Methods that include...

21. The method according to claim 20, wherein the substance is at least one plastic flake such as PET flakes, PE flakes, PO flakes, and / or PVC flakes, or at least one metal flake such as aluminum or an alloy thereof, or comprises the same.

22. The method according to claim 20 or 21, wherein the step of causing the automatic sorting machine to accept the substance includes refraining from transmitting a discharge signal to the discharge device (104) of the automatic sorting machine for the substance (12'), and the step of causing the automatic sorting machine to reject the substance includes transmitting a discharge signal to the discharge device (104) for the substance (12''').

Images