Plastic recycling support device and plastic recycling support method

Abstract

To enable accurate estimation of blending of an additive for recycling even waste plastic having unknown use history into recycled plastic having desired physical properties.SOLUTION: There is provided equipment 100 for supporting plastic recycling in which plastic is recycled into a recycled plastic having desired physical properties by blending an additive into the plastic, the equipment comprises: a physical property and deterioration estimator 140 which estimates physical properties and degree of deterioration of the plastic from a feature quantity of organization structure, extracted from surface analysis data of the plastic: and a blending estimator 150 which estimates physical properties of the recycled plastic from the physical properties of the plastic and degree of physical property deterioration, and blending conditions of the additive using a physical property recovery model.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a plastic recycling support device and a plastic recycling support method. [Background technology]

[0002] From the perspective of efficient resource utilization and reducing CO2 emissions, there is a need to improve the recycling rate of plastics.

[0003] Patent Document 1 discloses that in order to expand the scope of recovered materials used in the production of recycled thermoplastic resins, a compounding composition of additional materials to be added to recovered materials is calculated based on an index indicating the state of the recovered materials and a target value for resin design. A relational expression prepared in advance is used for this calculation. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2002-308998 Summary of the Invention [Problem to be solved by the invention]

[0005] Typically, the physical properties of plastics can be controlled by adding additives. However, in the case of recycled plastics, the physical properties of the waste plastics used as raw materials are unknown, making it difficult to determine the optimal additive blend to obtain recycled plastics with the desired properties. Furthermore, since waste plastics are often degraded due to oxidation or thermal history, it is necessary to understand the degree of degradation of the waste plastics and optimize the additive blend depending on the degradation, but the method for doing so is unclear.

[0006] The object of the present invention is to make it possible to determine, based on data, whether waste plastics with unknown usage histories can be recycled into recycled plastics with desired physical properties, and to make it possible to accurately estimate the blend of additives required to recycle waste plastics with unknown usage histories into recycled plastics with desired physical properties. [Means for solving the problem]

[0007] One embodiment of the plastic recycling support device of the present invention is a plastic recycling support device that supports plastic recycling by blending additives into plastic to recycle it into recycled plastic with desired physical properties, and has: a property / deterioration estimator that uses a property / deterioration estimation model to estimate the physical properties and degree of deterioration of the plastic from tissue structure features extracted from surface analysis data of the plastic; and a blending estimator that uses a property recovery model to estimate the physical properties of the recycled plastic from the physical properties, degree of deterioration of the plastic, and blending conditions of the additives, where the property / deterioration estimator estimates the physical properties and degree of deterioration of the sample from the tissue structure features extracted from the surface analysis data of the sample, and the blending estimator inversely estimates the blending conditions of the additives to be blended into the sample from the physical properties and degree of deterioration of the sample estimated by the property / deterioration estimator and the desired physical properties of the recycled plastic. [Effects of the Invention]

[0008] The invention provides a highly reliable plastic recycling process. By expanding the range of usable waste plastics, it leads to an improvement in the recycling rate. Other objects and novel features will become apparent from the description of this specification and the accompanying drawings. [Brief explanation of the drawings]

[0009] [Figure 1A] FIG. 1 is a diagram for explaining a plastic recycling support scheme. [Figure 1B] 10 shows examples of surface analysis methods, physical properties, deterioration levels, and additives for extracting structural features. [Figure 2] 1 is an example of an XRD spectrum. [Figure 3] FIG. 1 is a diagram showing a plastic recycling support system. [Figure 4] 1 is a configuration example of an information processing device. [Figure 5] 10 is a flowchart showing the entire plastic recycling support process. [Figure 6] 10 is a flowchart for determining acceptance criteria. [Figure 7] 10 is a flowchart showing the details of the candidate model construction step. [Figure 8] This is an example of a display screen presenting candidate models for property and deterioration estimation models. [Figure 9A] 10 is a flowchart showing details of an acceptance determination step. [Figure 9B] FIG. 10 is a diagram schematically illustrating a tissue structure feature space of a model. [Figure 10A] 10 is a flowchart showing details of an acceptance determination step. [Figure 10B] FIG. 10 is a diagram schematically illustrating a tissue structure feature space of a model. [Figure 11] 10 is an example of an input screen. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0011] The plastic recycling support scheme of this embodiment is shown in Figure 1A. In this embodiment, surface analysis data of the waste plastic used as the base material is used to extract multiple features (hereinafter referred to as "structural features") regarding the microstructural structure of the base material, and the physical properties and degree of degradation of the base material are estimated from the extracted microstructural features (first estimation). A property / degradation estimation model is used for the first estimation. The property / degradation estimation model uses the microstructural features of the base material as explanatory variables and the physical properties and degree of degradation of the base material as objective variables. Below, an example is described in which a model trained using machine learning techniques is used as the property / degradation estimation model. Next, based on the estimated physical properties and degree of degradation of the base material, the compounding conditions for additives to be blended into the waste plastic (base material) to obtain a recycled plastic (compound) with desired physical properties are estimated (second estimation). A property recovery model is used for the second estimation. The property recovery model is a model that uses the properties of the base material, the degree of deterioration, and the additive compounding conditions as explanatory variables, and the physical properties of the compound as the objective variable. Below, we will explain an example of using a model trained using machine learning techniques as the property recovery model.

[0012] FIG. 1B shows examples of surface analysis methods, physical properties, degradation levels, and additives for extracting structural features used in the first estimation. These are merely examples, and the present invention is not limited to these. Note that no generalized evaluation or measurement method for plastic degradation is known. As an example, an accelerated aging test is performed on plastic, and the degradation level is determined based on the conditions of the accelerated aging test. For example, the degradation level can be quantitatively determined so that the degradation level of plastics subjected to a longer accelerated aging test is greater.

[0013] Here, an example of extracting structural features of a base material from surface analysis data will be described. Figure 2 shows an XRD spectrum obtained by performing X-ray diffraction on the base material. The horizontal axis represents the diffraction angle, and the vertical axis represents the diffracted X-ray intensity. For example, a pseudo-Voigt function shown in (Equation 1) is fitted to such an XRD spectrum as a fitting function.

[0014]

number

[0015] By fitting the pseudo-Voigt function, four structural features (Δ2θ0: peak position, A: peak height, H k The spectrum example in Figure 2 contains 85 peaks, and these four tissue structure features are extracted for each peak.

[0016] FIG. 3 shows a plastic recycling support system. The plastic recycling support system includes a plastic recycling support device 100, which is communicatively connected to a terminal 210 via a network 200. The terminal 210 includes a display device 211 such as a display and an input device 212 such as a keyboard. A user accesses the plastic recycling support device 100 through the terminal 210 and, using the scheme shown in FIG. 1A, determines whether a waste plastic can be used to obtain a compound with the desired physical properties. If possible, the user determines the blending conditions for additives to be added to the waste plastic (base material). Surface analysis data and physical property data of the waste plastic are transmitted from the terminal 210 to the plastic recycling support device 100. FIG. 3 shows surface analysis devices such as a differential scanning calorimeter (DSC) 221, a Fourier transform infrared spectrophotometer (FTIR) 222, and an X-ray diffractometer (XRD) 223, and physical property measurement devices such as an impact resistance measurement device 224 and a melt mass-flow rate (MFR) measurement device 225. These are merely examples, and the present system is not limited to these devices.

[0017] The plastic recycling support device 100 is realized by an information processing device including, as shown in FIG. 4, a processor (CPU) 11, a memory 12, a storage device 13, an input device 14, an output device 15, a communication device 16, and a bus 17 as its main components. The processor 11 functions as a functional unit (functional block) that provides a predetermined function by executing processing according to a program loaded into the memory 12. The storage device 13 stores data and programs used by the functional unit. A non-volatile storage medium such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive) is used for the storage device 13. The input device 14 is a keyboard, a pointing device, etc., and the output device 15 is a display, etc. The communication device 16 enables communication with a terminal 210 or other information processing devices via a network 200. These are connected to each other via a bus 17 so that they can communicate with each other.

[0018] The plastic recycling support device 100 does not have to be realized by a single information processing device, but may be realized by multiple information processing devices. Also, some or all of the functions of the plastic recycling support device 100 may be realized as a cloud application.

[0019] The processing of the plastic recycling support device 100 will be described below with reference to the flowchart and the functional block diagram of the plastic recycling support device 100 shown in FIG.

[0020] FIG. 5 is a flowchart illustrating the overall plastic recycling support process. The user inputs the type of waste plastic to be recycled and the target specifications for the recycled plastic (S01). FIG. 11 shows an example of an input screen displayed on the terminal 210. The input screen 500 includes a waste plastic information input section 501 for the waste plastic to be recycled, a target specification input section 502, and prediction condition input sections 503 to 505. The information input through the waste plastic information input section 501 includes the type of waste plastic that will be used as the base material for the recycled plastic. Plastics include polypropylene (PP), polyethylene (PE), polystyrene (PS), and blends thereof, and recycling is performed based on the type of base material. It is also desirable to input information about the source of the waste plastic. The target specifications for the recycled plastic are input through the target specification input section 502. The target specifications include the name, target value, and tolerance range of a physical property parameter. Note that, hereinafter, a physical property parameter defined as a target specification is referred to as a target physical property parameter, and unless otherwise specified, a target value including a tolerance range is referred to as a target physical property parameter value. The number of physical property parameters that can be used as target specifications is not limited. Furthermore, the prediction condition input section allows the user to input in advance conditions for selecting a model to be used in the plastic recycling support process, making it easier to narrow down the models by the plastic recycling support device 100. Here, an example is shown in which the user inputs the estimation accuracy 503 of the model, the cost 504 and time 505 allowed for surface analysis to obtain input data that will be explanatory variables for the property / deterioration estimation model.

[0021] The user obtains a sample of the waste plastic to be recycled (S02), and the plastic recycling support device 100 performs an acceptance decision on the sample (S03). The sample is the type of waste plastic input in the input step S01, but the user does not have information about the sample's physical properties or degree of deterioration, and therefore does not know whether the sample can be recycled into a plastic (compound) with the desired characteristics. For example, if the physical properties of the base material significantly deviate from the target specifications for the recycled plastic or if the sample has deteriorated significantly, the target specifications may not be achieved. Therefore, in the acceptance decision step (S03), it is determined whether the sample is likely to satisfy the target physical property parameter values ​​input in the input step S01, and if it is determined that the target physical property parameter values ​​can be satisfied, the sample is deemed acceptable. Details of the acceptance decision step (S03) will be described later.

[0022] For the waste plastics that have been accepted, the plastic recycling support device 100 performs blend optimization (S04). In the blend optimization step (S04), since the physical properties and degree of deterioration of the sample have been estimated by the estimator 141 of the physical property / deterioration estimator 140, the inverse estimator 152 of the blend estimator 150 uses a property recovery model to estimate blending conditions for additives that satisfy the target physical property parameter values.

[0023] FIG. 6 is a flowchart for determining the acceptance criteria in the acceptance determination step (S03). This flow is mainly executed by the model selector 160. The model selector 160 first searches the model database 163 (S11) based on the type of waste plastic and the target specifications of the recycled plastic input in the input step (S01). The model database 163 stores models previously created by the plastic recycling support device 100. In the case of a model based on machine learning, whether appropriate inference is possible depends on the learning data used to train the model. Therefore, if a trained model that can be used for the input content in the input step (S01) is stored in the model database 163, the searcher 161 selects such a model as a candidate model (S12); if not, the searcher 161 constructs a candidate model (S13). If prediction conditions have been input by the user (see FIG. 11), the searcher 161 selects a model that satisfies those prediction conditions.

[0024] Details of the candidate model construction step (S13) are shown in Figure 7. Data stored in the plastic database 170 is referenced based on the type of plastic and the target physical property parameter values ​​(S21). The plastic DB 170 stores surface analysis data, physical property data, and degradation data for each type of plastic. Data on plastics in which the same sample has been subjected to accelerated degradation tests under different conditions is stored, and the degradation data is based on the conditions of this accelerated degradation test. Furthermore, physical property data on recycled plastics in which additives have been blended into these plastics and recycled, as well as the additive blending conditions at that time, are also stored.

[0025] A physical property / deterioration estimation model and a physical property recovery model are constructed using the data stored in the plastic DB 170 as training data (S22, S23). In the case of the physical property / deterioration estimation model, the learning device 111 of the first model constructor 110 constructs the model by performing supervised learning using, for example, a combination of the structural features of the base material (plastic subjected to an accelerated degradation test) and the physical properties and degradation level of the base material stored in the plastic DB 170 as training data. In the case of the physical property recovery model, the learning device 121 of the second model constructor 120 constructs the model by performing supervised learning using, for example, a combination of the physical properties, degradation level, and additive blending conditions of the base material and the physical properties of the compound (recycled plastic obtained by blending additives into the base material under the blending conditions and recycling it) stored in the plastic DB 170 as training data.

[0026] Here, it is desirable to construct multiple models for property / degradation estimation and property recovery. Generally, using multiple explanatory variables can improve the estimation accuracy of a model, but if multiple types of surface analysis are required, the cost and time required to obtain the analytical data will increase. Furthermore, the degree to which explanatory variables contribute to improving estimation accuracy will vary. For this reason, it is desirable to construct multiple models with different surface analysis methods for obtaining structural features and different physical property parameters to predict, allowing the user to select the optimal model by comparing and considering the accuracy of the model with the cost of obtaining the data required to use the model.

[0027] Therefore, the estimation accuracy and data acquisition cost are calculated for each constructed model (S24), and the constructed model is registered in the model database 163 in association with the type of plastic, target physical property parameter value, estimation accuracy, and data acquisition cost (S25). The calculation of the estimation accuracy of the model is performed by the accuracy calculator 112 of the first model constructor 110 and the accuracy calculator 122 of the second model constructor 120, respectively. Furthermore, when prediction conditions are input by the user (see FIG. 11), the first model constructor 110 and the second model constructor 120 construct models that satisfy those prediction conditions.

[0028] Returning to the explanation of Fig. 6, the model selector 160 presents the estimation accuracy and data acquisition cost for the selected or constructed candidate model to the terminal 210 (S14). Fig. 8 shows an example of a display screen 300 presenting candidate models for property and degradation estimation models.

[0029] Each model 301 corresponds to a candidate model, and for each candidate model, the number of parameters 302, surface analysis method 303, time 304, measurement cost 305, property estimation accuracy 306, and degradation estimation accuracy 307 are displayed. The number of parameters 302 is the number of parameters (in this case, tissue structure feature quantities) that serve as input data for the model. The surface analysis method 303 is a surface analysis method required to obtain the parameters (tissue structure feature quantities) that serve as input data. Multiple types of surface analysis may be required depending on the tissue structure feature quantities that serve as input to the model. The time 304 and measurement cost 305 indicate the time and cost required to obtain input data for the model using the method specified in the surface analysis method 303, respectively. The property estimation accuracy 306 and degradation estimation accuracy 307 indicate the estimation accuracy of the physical properties of the base material and the estimation accuracy of the degree of degradation of the base material, respectively, among the output data of the model. The estimation accuracy can be expressed, for example, as the coefficient of determination R 2 can be used.

[0030] The user selects the model to be used based on the model information presented on the terminal 210. This determines the surface analysis method for the base material (waste plastic) and the texture feature values ​​to be used in the analysis (S15).

[0031] The inverse estimator 152 of the blend estimator 150 uses the selected property recovery model to estimate the allowable range of the base material's physical properties and degradation level from the target physical property parameter values ​​input in the input step (S01) (S16). Next, the inverse estimator 142 of the property / degradation estimator 140 uses the selected property / degradation estimation model to convert the allowable range of the base material's physical properties and degradation level obtained in step S16 into a microstructure feature space and store it in the allowable range memory 162 (S17). In the flowchart, it is desirable to obtain not only the allowable range of the base material's physical properties and degradation level in step S16, but also the unacceptable range of the base material's physical properties and degradation level, and then convert each of these into a microstructure feature space in step S17. By doing so, as described below, it becomes possible to determine from the sample's microstructure feature whether the model can appropriately determine the acceptability of the base material's physical properties based on the physical properties of the sample, or whether the model cannot appropriately determine the acceptability of the sample.

[0032] The microstructure feature space indicating the allowable range of the base material's physical properties and degree of deterioration obtained in step S17 is the acceptance criterion used in the acceptance decision step (S03) (see Fig. 5). That is, in the acceptance decision step (S03), if the analysis data analyzed by the surface analysis method determined in step S15 is included in the microstructure feature space obtained in step S17, it is highly likely that the material can be regenerated to meet the target specifications input in input step S01, and it is determined to be acceptable.

[0033] A detailed example of the sample acceptance determination step (S03) is shown in FIG. 9A. First, the user performs a surface analysis of the sample using the surface analysis method determined in step S15 (S31). The surface analysis data is input from terminal 210 to data input unit 182 of plastic recycling support device 100. Feature extraction unit 181 receives the surface analysis data from data input unit 182 and extracts tissue structure features by fitting (S32). Thereafter, estimator 141 of physical property / deterioration estimator 140 inputs the tissue structure features, which are the input data, into a physical property / deterioration estimation model and estimates the physical properties and degree of deterioration of the sample (base material) (S33).

[0034] Here, we take as an example a model for estimating physical properties and deterioration, in which the input data is a first microstructure feature obtained from analytical data obtained by a first surface analysis method and a second microstructure feature obtained from analytical data obtained by a second surface analysis method. In reality, multiple microstructure feature values ​​can usually be extracted from analytical data obtained by a single surface analysis method, as shown in Figure 2, but for simplicity's sake, we will assume that one microstructure feature value is obtained from one surface analysis method. Figure 9B shows a schematic diagram of the microstructure feature space of the model.

[0035] In this example, the tissue structure feature space of the model is defined as a space 400 spanned by the first tissue structure feature and the second tissue structure feature. Region 401 is a region where the model can be estimated to be able to accept the sample, region 402 is a region where the model can be estimated to be unable to accept the sample, and the remaining region 403 is a region where the model cannot determine whether the sample is acceptable or unacceptable. For example, in a region where the model has not been trained using training data, the reliability of the model's inference decreases. Such a region with low reliability of inference is region 403. The ranges of these regions are stored in the tolerance range memory 162 of the model selector 160. The comparator 131 of the determiner 130 determines in which region of the tissue structure feature space of the model the tissue structure feature obtained from the analysis data is located (S34, S35).

[0036] Comparator 131 determines that the tissue structure feature obtained from the analysis data is acceptable when it is located in region 401 (S36), determines that the tissue structure feature obtained from the analysis data is unacceptable when it is located in region 402 (S37), and determines that the tissue structure feature obtained from the analysis data is not estimable by the model when it is located in region 403.

[0037] If comparator 131 determines that estimation is impossible, the user measures the physical property values ​​of the sample and stores the results in plastic database 170 (S38). Comparator 131 determines that the measured physical property values ​​are acceptable if they are within the allowable range of the physical property of the base material determined in step S16 (S36), and determines that the measured physical property values ​​are unacceptable if they are outside the allowable range of the physical property of the base material determined in step S16 (S40). Note that by storing the measured physical property values ​​of the sample together with the surface analysis data in plastic database 170, they can be used for subsequent model learning.

[0038] Another detailed example of the sample acceptance decision step (S03) is shown in FIG. 10A. The example in FIG. 9A illustrates a physical property / deterioration estimation model that uses as input data a first tissue structure feature obtained from analytical data obtained by a first surface analysis method and a second tissue structure feature obtained from analytical data obtained by a second surface analysis method. In contrast, the flowchart in FIG. 10A first performs an acceptance decision using a first physical property / deterioration estimation model that uses as input data the first tissue structure feature obtained from analytical data obtained by the first surface analysis method. If the first physical property / deterioration estimation model is unable to make a decision, the second physical property / deterioration estimation model that uses as input data the first tissue structure feature and the second tissue structure feature obtained from analytical data obtained by the second surface analysis method. This allows for a lower-cost acceptance decision when the first physical property / deterioration estimation model can determine whether a sample is acceptable.

[0039] A user first performs a surface analysis of a sample using a first surface analysis method (S51). Surface analysis data is input from terminal 210 to data input unit 182 of plastic recycling support device 100. Feature extraction unit 181 receives the surface analysis data from data input unit 182 and extracts first tissue structure feature values ​​by fitting (S52). Then, estimator 141 of physical property / deterioration estimator 140 inputs the first tissue structure feature values, which are input data, into a first physical property / deterioration estimation model, and estimates the physical properties and degree of deterioration of the sample (base material) (S53).

[0040] Figure 10B shows a schematic diagram of the tissue structure feature space of the first physical property / deterioration estimation model and the second physical property / deterioration estimation model. The tissue structure feature space of the second physical property / deterioration estimation model is the same as the tissue structure feature space shown in Figure 9B. Meanwhile, the tissue structure feature space of the first physical property / deterioration estimation model is a one-dimensional space divided into regions 411 to 414. Region 411 is a region that can be determined as acceptable by the first physical property / deterioration estimation model. Region 412 is a region that can be determined as unacceptable by the first physical property / deterioration estimation model. Regions 413 and 414 are regions that cannot be determined as acceptable by the first physical property / deterioration estimation model. Region 413 is a region where acceptability changes depending on the second tissue structure feature, so the reliability of the inference results of the first physical property / deterioration estimation model should be low. Region 414 is a region where the model has not been trained using training data, so the reliability of the inference results is also low in this region, making it impossible to determine.

[0041] Comparator 131 of determiner 130 determines in which region of the one-dimensional tissue structure feature space of the first physical property / deterioration estimation model the tissue structure feature obtained from the analysis data is located (S54, S55). Comparator 131 determines that the first tissue structure feature obtained from the analysis data is acceptable if it is located in region 411 (S56), determines that the first tissue structure feature is unacceptable if it is located in region 412 (S57), and determines that the first tissue structure feature is indeterminable by the first physical property / deterioration estimation model if it is located in regions 413 or 414.

[0042] If the comparator 131 determines that the determination is impossible, the comparator 131 switches the model used from the first physical property / deterioration estimation model to the second physical property / deterioration estimation model (S58). The user performs a surface analysis of the sample using the second surface analysis method (S59). The surface analysis data is input from the terminal 210 to the data input unit 182 of the plastic recycling support device 100. The feature extraction unit 181 receives the surface analysis data from the data input unit 182 and extracts second tissue structure features by fitting (S60). Thereafter, the estimator 141 of the physical property / deterioration estimator 140 inputs the first tissue structure feature and the second tissue structure feature, which are the input data, into the second physical property / deterioration estimation model and estimates the physical properties and degree of deterioration of the sample (base material) (S61). The processing from step S61 onwards is the same as the processing from step S33 onwards in the flowchart shown in FIG. 9A, and therefore a redundant description will be omitted.

[0043] The present invention is not limited to the above-described embodiments, but includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those including all of the described configurations. Furthermore, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace part of the configuration of each embodiment with other configurations. [Explanation of symbols]

[0044] 11: Processor (CPU), 12: Memory, 13: Storage device, 14: Input device, 15: Output device, 16: Communication device, 17: Bus, 100: Plastic recycling support device, 110: First model builder, 111: Learning device, 112: Accuracy calculator, 120: Second model builder, 121: Learning device, 122: Accuracy calculator, 130: Decision device, 131: Comparator, 140: Physical property / deterioration estimator, 141: Estimator, 142: Inverse estimator, 150: Blend estimator, 151: Estimator, 152: Inverse estimator, 160: Model selector, 161: Searcher, 162: Tolerance range memory, 163: Model database, 170: Plastic database, 181: Feature extraction unit, 182 : Data input unit, 200: Network, 210: Terminal, 211: Display device, 212: Input device, 221: Differential scanning calorimeter, 222: Fourier transform infrared spectrophotometer, 223: X-ray diffraction device, 224: Impact resistance measurement device, 225: Melt mass-flow rate measurement device, 300: Display screen example, 301: Model, 302: Number of parameters, 303: Surface analysis method, 204: Time, 305: Measurement cost, 306: Physical property estimation accuracy, 307: Deterioration estimation accuracy, 400: Space, 401, 402, 403, 411, 412, 413, 414: Area, 500: Input screen, 501: Waste plastic information input unit, 502: Target specification input unit, 503, 504, 505: Prediction condition input unit.

Claims

1. A plastic recycling support device that supports plastic recycling by blending additives into plastic to recycle it into recycled plastic having desired physical properties, a property / degradation estimator that estimates the property and degree of degradation of a plastic from tissue structure features extracted from surface analysis data of the plastic using a property / degradation estimation model; A blending estimator that uses a property recovery model to estimate the physical properties of recycled plastics from the physical properties of the plastics, the degree of deterioration, and the blending conditions of additives, the physical property / deterioration estimator estimates the physical property and deterioration level of the sample from a tissue structure feature extracted from surface analysis data of the sample; The blending estimator is a plastic recycling support device that reverse-estimates the blending conditions of additives to be blended into the sample based on the physical properties and degree of deterioration of the sample estimated by the physical property / deterioration estimator and the desired physical properties of the recycled plastic.

2. In claim 1, a determiner for determining whether the sample is acceptable; the determiner determines whether the sample is acceptable or not based on whether or not a tissue structure feature extracted from the surface analysis data of the sample is within an allowable range according to desired physical properties of the recycled plastic; The plastic recycling support device determines the allowable range by having the blend estimator inversely estimate the allowable range of physical properties and degradation level allowed for the sample from the desired physical properties of the recycled plastic, and the physical property / degradation estimator converts the allowable range of physical properties and degradation level into a tissue structure feature space.

3. In claim 2, a model database storing a plurality of the physical property / deterioration estimation models; The physical property / deterioration estimation model is a plastic recycling support device that stores the surface analysis method, time, and measurement cost required to extract the tissue structure feature quantities to be input into the model, in association with the physical property estimation accuracy and the deterioration estimation accuracy.

4. In claim 2, The plastic recycling support device, wherein the judger judges the sample to be acceptable if the tissue structure feature extracted from the surface analysis data of the sample is a value in a range that cannot be estimated by the physical property / deterioration estimation model, and if the physical properties measured for the sample satisfy the allowable range of physical properties for the sample that the blend estimator has determined by inverse estimation from the desired physical properties of the recycled plastic.

5. In claim 2, The determiner determines whether the sample is acceptable based on whether the tissue structure feature extracted from the surface analysis data of the sample is within a first tolerance range corresponding to the desired physical properties of the recycled plastic, and if the sample cannot be accepted based on the first tolerance range, determines whether the sample is acceptable based on whether the tissue structure feature is within a second tolerance range corresponding to the desired physical properties of the recycled plastic; the first allowable range is determined by the physical property / deterioration estimator converting the allowable ranges of the physical property and the degree of deterioration into a tissue structure feature space using a first physical property / deterioration estimation model, and the second allowable range is determined by the physical property / deterioration estimator converting the allowable ranges of the physical property and the degree of deterioration into a tissue structure feature space using a second physical property / deterioration estimation model; A plastic recycling support device in which the first physical property / deterioration estimation model requires fewer types of surface analysis methods to extract tissue structure features to be input to the model than the second physical property / deterioration estimation model.

6. In claim 5, The plastic recycling support device, wherein the judger judges the sample to be acceptable if the tissue structure feature extracted from the surface analysis data of the sample is a value in a range that cannot be estimated by the second physical property / deterioration estimation model, and if the physical properties measured for the sample satisfy the allowable range of physical properties for the sample that the blend estimator has determined by inverse estimation from the desired physical properties of the recycled plastic.

7. In claim 3, a plastic database that stores, for each type of plastic, surface analysis data, physical property data, and degradation degree data based on the conditions of an accelerated degradation test performed on the plastic; A plastic recycling support device having a first model builder that builds the physical property / deterioration estimation model using a combination of tissue structure features extracted from the surface analysis data of the plastic stored in the plastic database and physical property data and degradation degree data of the plastic as learning data.

8. In claim 7, a second model builder; The plastic database stores, for each type of plastic, data on the physical properties of recycled plastics obtained by blending additives into plastics after an accelerated degradation test, and the blending conditions of the additives, The second model builder is a plastic recycling support device that builds the property recovery model using a combination of the physical property data of the plastic, deterioration degree data, and additive compounding conditions stored in the plastic database and the physical property data of the recycled plastic as learning data.

9. A plastic recycling support method using a plastic recycling support device that supports plastic recycling by blending additives into plastic to recycle it into recycled plastic having desired physical properties, The plastic recycling support device includes a property / deterioration estimator that uses a property / deterioration estimation model to estimate the property and degree of deterioration of a plastic from a tissue structure feature extracted from surface analysis data of the plastic, and a blending estimator that uses a property recovery model to estimate the property of a recycled plastic from the property, degree of deterioration, and blending conditions of an additive; the physical property / deterioration estimator estimates the physical property and deterioration level of the sample from a tissue structure feature extracted from surface analysis data of the sample; A plastic recycling support method in which the blending estimator reverse-estimates the blending conditions of additives to be blended into the sample based on the physical properties and degree of deterioration of the sample estimated by the physical property / deterioration estimator and the desired physical properties of the recycled plastic.

10. In claim 9, The plastic recycling support device has a determiner that determines whether the sample is acceptable, the determiner determines whether the sample is acceptable or not based on whether or not a tissue structure feature extracted from the surface analysis data of the sample is within an allowable range according to desired physical properties of the recycled plastic; A plastic recycling support method in which the allowable range is determined by the blend estimator inversely estimating the allowable ranges of physical properties and degradation levels permitted for the sample from the desired physical properties of the recycled plastic, and the physical property / degradation estimator converting the allowable ranges of physical properties and degradation levels into a tissue structure feature space.

11. In claim 10, The plastic recycling support device has a model database that stores a plurality of the physical property / deterioration estimation models, A plastic recycling support method in which the physical property / deterioration estimation model stores the surface analysis method, time and measurement cost required to extract the tissue structure features to be input into the model, as well as the physical property estimation accuracy and the deterioration estimation accuracy in association with each other.

12. In claim 10, In a plastic recycling support method, when the tissue structure feature extracted from the surface analysis data of the sample is a value in a range that cannot be estimated by the physical property / deterioration estimation model, the determiner determines that the sample is acceptable if the physical properties measured for the sample satisfy the allowable range of physical properties that are allowable for the sample, which are determined by the blend estimator by inverse estimation from the desired physical properties of the recycled plastic.

13. In claim 11, The plastic recycling support device includes a plastic database that stores, for each type of plastic, surface analysis data, physical property data, and degradation degree data based on the conditions of an accelerated degradation test performed on the plastic, and a first model builder; A plastic recycling support method in which the first model builder builds the physical property / deterioration estimation model using as learning data a combination of tissue structure features extracted from the surface analysis data of the plastic stored in the plastic database and physical property data and deterioration degree data of the plastic.

14. In claim 13, The plastic recycling support device includes a second model builder, The plastic database stores, for each type of plastic, data on the physical properties of recycled plastics obtained by blending additives into plastics after an accelerated degradation test, and the blending conditions of the additives, A plastic recycling support method in which the second model builder constructs the property recovery model using a combination of the physical property data of the plastic, deterioration level data, and additive compounding conditions stored in the plastic database and the physical property data of the recycled plastic as learning data.

Images