JP2001091484A - Plastic mixture combustion calorie measuring method and device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a plastic mixture combustion calorie measuring method and device capable of measuring the combustion calorie under the plastic mixture state of mixing a plurality of plastic pieces. SOLUTION: An absorbance spectrum of a plastic mixture, or a measurement object is measured (S1). Assuming that the absorbance spectrum is linear sum of the absorbance spectra as a material simple substance, its coefficient is calculated by a multiple regression analysis (S2). Then, assuming that the obtained each coefficient value is the volume of the each component material, each weight is calculated by multiplied it by the density of the each component material (S3) and the obtained weight is multiplied by the combustion calorie per unit weight of each component material so as to calculate the combustion calorie of the plastic mixture (S4).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for measuring the calorie of combustion of waste plastic used as a raw material for combustion, and more particularly to a method of measuring the calorie of combustion in the state of a plastic mixture in which plural kinds of plastic pieces are mixed. TECHNICAL FIELD The present invention relates to a method and an apparatus for measuring the calorie of combustion of a plastic mixture capable of measuring the calorific value.

[0002]

2. Description of the Related Art When waste plastics are used as a raw material for combustion, in order to stabilize the combustion state in the furnace, the combustion calories of the waste plastics to be injected are grasped, and the amount and the number of times of injection are controlled based on this. There is a need to. Here, as a method for stabilizing the combustion state in a furnace when waste plastic is used as a raw material for combustion, Japanese Patent Application Laid-Open No. Hei 6-21063 discloses a method.
No. 2 proposes a method in which waste plastics are separated by type (combustion calories) before the waste plastics are put into a furnace, and only necessary plastics are crushed and put into a furnace. In addition, the above separation method involves irradiating the waste plastic with infrared light or near-infrared light to detect its absorption characteristics, and comparing the absorption characteristics of each plastic in the same wavelength region based on the molecular structure of the waste plastics. It is to determine the type of plastic or the calories burned.

[0003]

However, according to the above-mentioned prior art, only the required amount of plastic is put into a furnace after separating waste plastic for each type. In addition, there is a problem that a place for storing each separated plastic is required for each type. Further, when a plurality of types of plastics are mixed in the measurement target range in one measurement, there is also a problem that the calorie of combustion cannot be measured. Since waste plastics are usually collected in a state where various kinds of plastics are mixed in a random manner, it is actually difficult to perform the above-described measurement. by the way,
When waste plastic is used as fuel in a small combustion furnace or a blast furnace, it is desirable that the waste plastic be crushed or reduced in size and then introduced in order to simplify the charging device and stabilize the combustion state. In addition, waste plastics are often collected in a form in which a plurality of different types of plastics are mixed. For this reason, the recovered waste plastic is crushed and crushed as it is to produce a waste plastic mixture in which multiple types of plastics are mixed at an unknown ratio, and the combustion calories of the plastic mixture are measured before being injected. If possible, it is possible to solve the problems in the above conventional method. The present invention has been made in view of the above circumstances, and its purpose is to
An object of the present invention is to provide a method and an apparatus for measuring the calorie of combustion of a plastic mixture, which can measure the calorie of combustion in a state of a plastic mixture in which a plurality of types of plastic pieces are mixed.

[0004]

In order to achieve the above object, a method of the present invention is a method for measuring the calorie burned of a plastic mixture in which a plurality of types of plastic pieces are mixed. A mixture spectrum measuring step for measuring the absorbance spectrum of the plastic mixture crushed or reduced to a size smaller than the size, a plurality of reference samples having known absorbance spectra and a calorific value of the plastic mixture obtained in the mixture spectrum measuring step A combustion calorie calculating step of calculating the calorie burned of the plastic mixture based on a predetermined evaluation function based on the absorbance spectrum of the plastic mixture. Have been. This makes it possible to crush and recycle the collected waste plastic as it is, to produce a waste plastic mixture in which multiple types of plastics are mixed at an unknown ratio, and to measure the combustion calories of the plastic mixture before putting it in. Therefore, there is no need to perform processing such as sorting waste plastics by type before and after measurement, and there is no problem in the storage location of each separated plastic.
As described above, it is possible to rationalize the apparatus and the processing procedure as compared with the related art. Here, if the plastic mixture has already been atomized, the above method can be carried out using the plastic mixture as it is. However, if not, the plastic mixture is previously crushed or reduced to a predetermined size or smaller. A process is required.

In the calorific value calculation step, the evaluation function is obtained by calculating a linear sum of spectra obtained by weighting the absorbance spectra of the plastic mixture as the reference sample and the absorbance spectra of the plural kinds of plastics alone. The weight is calculated by the multiple regression method by applying the absorbance spectrum of the plastic mixture obtained in the mixture spectrum measuring step to the approximate regression equation, and the plastic mixture is formed based on the weight. Calculating the volume ratio of each plastic, and further calculating the calorie burned of the plastic mixture based on the volume ratio, the density of the plurality of types of plastics alone, and the calorie burned of the plurality of types of plastics alone. It is possible to do. Alternatively, the evaluation function may be configured as a regression equation obtained by a principal component regression method or a partial least squares method based on the absorbance spectra of a plurality of plastic mixtures whose calorie burned as the reference sample is known. Furthermore, if an absorbance spectrum correction step is provided to remove the influence of the particle size distribution of the plastic mixture from the absorbance spectrum of the plastic mixture obtained in the mixture spectrum measurement step, there is some variation in the particle diameter of the plastic mixture. Even if it is corrected, high-precision measurement becomes possible. As the processing in the absorbance spectrum correction step, a method using a second derivative operation of the absorbance spectrum or a method of comparing the absorbance spectrum obtained from a predetermined reference plastic with the absorbance spectrum of the plastic mixture obtained in the mixture spectrum measurement step is used. A method of correcting the absorbance spectrum of the above-mentioned plastic mixture so as to minimize the sum of squares of the difference, that is, a method using a so-called MSC process, or the like can be considered. Further, if the absorbance spectrum of the plastic mixture obtained in the mixture spectrum measuring step is corrected so that at least one of the spatial spread and the temporal spread of the plastic mixture is adjusted to a predetermined value, the absorbance spectrum at the time of spectrum measurement is corrected. Even if there is a variation in the amount of the plastic mixture, it is possible to correct the variation and perform highly accurate measurement.

In order to achieve the above object, an apparatus according to the present invention is a plastic mixture combustion calorie measuring apparatus for measuring the combustion calories of a plastic mixture in which a plurality of types of plastic pieces are mixed. Alternatively, a mixture spectrum measuring means for measuring the absorbance spectrum of the small-sized plastic mixture, the absorbance spectrum of the plastic mixture obtained by the mixture spectrum measuring means, and the absorbance spectra of a plurality of reference samples having known calories burned. A combustion calorie calculating means for calculating the calorie burned of the plastic mixture based on a predetermined evaluation function based on the evaluation function. In addition, the above-mentioned combustion calorie measuring method can all be implemented on this apparatus.

[0007]

Embodiments and examples of the present invention will be described below with reference to the accompanying drawings to provide an understanding of the present invention. The following embodiments and examples are mere examples embodying the present invention, and do not limit the technical scope of the present invention. Here, FIG. 1 is a schematic diagram showing a schematic configuration of the combustion calorie measurement A1 according to the embodiment of the present invention, FIG. 2 is a flowchart showing an example of a combustion calorie calculation processing procedure when using a multiple regression method, and FIG. Is PP, PET, P
FIG. 4 shows the results of measurement of the absorbance spectrum of the simple substance S, FIG. 4 shows the results of the measurement of the absorbance spectrum of the mixture of two kinds (PP: PET = 1: 1), and FIG. 5 shows the result of measurement of the mixture of three kinds (PP: PET: PS = 1: 1).
1: 1) Absorbance spectrum measurement results, FIG. 6 shows the results of measurement of absorbance spectra in three different states of plate, pellet, and powder in PE, and FIG. 8 and FIG. 8 are explanatory diagrams of the principal component analysis, and FIG.
1 is a flowchart showing an example of a procedure for calculating calorie burned when the PCR method is used. FIG. 10 is a flowchart showing an example of a procedure for calculating calorie burned when the PLS method is used.
1 is a comparison diagram of the absorbance spectrum of an unknown sample before and after MSC treatment and the absorbance spectrum of a reference sample. FIG. 12 shows various plastic mixtures pulverized into powder.
The results of estimating the calories burned using the three regression methods,
FIGS. 13A and 13B show a case where pre-processing (secondary differentiation processing and MSC processing) is not performed, a case where MSC processing is performed, and a case where second-order differentiation processing is performed. FIG. 13 shows combustion according to an embodiment of the present invention. It is a schematic diagram which shows the schematic structure of calorie measurement A2.

[1. Apparatus Configuration] The combustion calorie measuring apparatus A1 according to the present embodiment has a schematic configuration as shown in FIG. Waste plastics 0 that have been recovered to be used as combustion raw materials are directly separated into crushers 1 without being separated by type.
And crushed together here. Then, the crushed particles crushed to a predetermined size (for example, about 10 mm) or less are sorted by the particle size sorter 2 and put into the mixer 3. Since the types of the crushed particles to be supplied to the mixer 3 are uneven to some extent, the crushed particles of each type are uniformed as a whole. The plastic mixture 0 'homogenized by the mixer 3 is spread on the conveyor 4 as thinly and uniformly as possible, and is flowed. Above the conveyor 4, a light source 11 and a spectroscope 12 are installed. The light source 1 irradiates the plastic mixture 0 'flowing on the conveyor 4 with white light, and the plastic mixture 0'. The spectrometer 12 spectroscopically measures the reflected light from the sample to determine the absorbance (= lo) of the plastic mixture 0 '.
g (1 / reflectance)) is obtained as a near-infrared absorption spectrum (hereinafter, referred to as an absorbance spectrum). Here, the light source 11 and the spectroscope 12 are an example of a mixture spectrum measuring unit. The absorbance spectrum obtained by the spectroscope 12 is sent to a calorie calculator 13, where it is used in a calorie calculation process described later.
The calorie calculator 13 also receives, as necessary, speed information of the conveyor 4 and information such as the amount of the plastic mixture 0 'sent by the conveyor 4, and uses the information in the calorie calculation process. When the calorie calculator 13 (an example of combustion calorie calculation means) obtains the combustion calories of the plastic mixture 0 'sent by the conveyor 4, the plastic mixture 0' is converted into a value of the combustion calories by a combustion controller (not shown). Is charged into the combustion furnace 6 via the conveyor 5 in accordance with the charging amount control based on.

[2. Combustion Calorie Calculation Process] Subsequently, the plastic mixture 0 'in the calorie calculator 13 is described.
The procedure of the combustion calorie calculation process will be described. In the present measurement apparatus A1, the absorbance spectrum of the plastic mixture 0 'is considered to be represented by a linear sum of the spectra obtained by weighting the absorbance spectra of each kind of plastic alone contained in the mixture 0'. The weight was calculated by a regression method, and the ratio of the weight was regarded as the volume ratio of each plastic contained in the mixture 0 ', and the calorie burned of the mixture 0' was estimated. Before describing the specific procedure of the above processing, the result of performing verification on the above processing will be described.

(A. Verification of Combustion Calorie Calculation Process) First, the absorbance spectra of three types of plastic raw materials, PP (polypropylene), PET (polyethylene terephthalate), and PS (polystyrene), were measured. In the above measurement, a sampling interval of 2 nm was set in a wavelength range of 1100 nm to 2500 nm. As the measurement results, those obtained by subjecting the obtained absorbance spectrum to second-order differentiation were used (this will be described later). FIG. 3 shows the measurement results of the individual raw materials. It can be seen that the obtained absorbance spectra differ greatly for each raw material.
As shown in Table 1 below, three kinds of the two kinds of mixtures and three kinds of the three kinds of the mixtures were used using the above three kinds of plastic raw materials.
Generated as follows. The mixing ratios in the table are weight ratios.

[Table 1] Then, the absorbance spectra of the various mixtures were measured five times for one sample, and the obtained absorbance spectra were secondarily differentiated in the same manner as in the case of the simple substance. FIG. 4 shows the measurement results (PP: PET = 1: 1) of the two kinds of mixture, and FIG. 5 shows the measurement results (PP: PET: PS = 1: 1: 1) of the three kinds of mixture. Here, the spectrum A (λ) of the mixture
Is considered to be a superposition of the spectra S _k (λ) of the M single raw materials constituting the mixture, and the following equation is assumed, with the weight of each single raw material being W _k (k = 1 to M). A (λ) = {W _k × S _k (λ)} (k = 1 to M) (1) Applying the absorbance spectrum of the single raw material and the absorbance spectrum of the mixture to the above equation (1) A multiple regression analysis was performed, and the values of the obtained coefficients W _k are shown in Table 2.

[Table 2] Further, when the coefficients in Table 2 are normalized so that the sum becomes 1, the following is obtained.

[Table 3]

Table 4 shows the actual volume ratio (normalized) of each single raw material in each mixture, and Table 5 shows the same weight ratio. That is, Table 5 is obtained by multiplying Table 4 by each density.

[Table 4]

[Table 5] Comparing Table 3 with Tables 4 and 5, it can be seen that the value of each coefficient (Table 3) obtained by multiple regression analysis is very close to the volume ratio of each single raw material (Table 4). Therefore, it is possible to estimate the volume of each single raw material by assuming each coefficient obtained by multiple regression analysis as the volume ratio of each single raw material, and obtain the weight of each single raw material by multiplying those values by the density. It is possible. Furthermore, by multiplying the weight of each single raw material by the value of the calorie burned per unit weight and summing up, the calorie burned of the plastic mixture can be estimated.

(B. Elimination of the influence on the measurement result by the difference in the particle size of the crushed material) Next, the influence on the measurement result by the difference in the particle size of the crushed material will be considered. FIG. 6 shows the results of measuring the absorbance spectrum by processing PE (polyethylene) into three different states: a plate, a pellet having a particle size of 3 to 5 mm, and a powder having a particle size of 1 mm or less. From these results, the absorbance spectrum obtained is offset due to the difference in the particle size of the crushed plastic, and if the calorie measurement as described above is performed using this as it is, the difference in the particle size greatly affects the measurement result. It is easy to guess. Therefore, the measuring device A1 solves the above-described problem by using the second derivative of the measured absorbance spectrum, as described above. FIG.
As shown in (1), the constant offset is removed by performing first-order differentiation on the raw data of the absorbance spectrum, and the linear offset is further removed by performing second-order differentiation. In this way, the second-order differentiation of the measured absorbance spectrum removes the offset due to the difference in particle size as shown in FIG. 6 and suppresses the influence on the measurement result due to the difference in particle size of the crushed material. It becomes possible.

(C. Specific Procedure of Combustion Calorie Calculation Process) Next, a specific procedure of the calculation process of combustion calories will be described with reference to the flowchart shown in FIG. First, before you start measuring plastic mixtures,
Absorbance spectra (second derivative) of all the raw materials constituting the waste plastic are collected and stored in a predetermined memory (step S0). This step S0
Need only be done once at the beginning unless new plastic material is added. When the measurement of the plastic mixture is started, the absorbance spectrum of the plastic mixture 0 'flowing on the conveyor 4 is measured by the spectroscope 12, and sent to the calorie calculator 13 (step S1).
In the calorie calculator 13, the absorbance spectrum of the plastic mixture 0 ′ obtained from the spectrometer 12 is secondarily differentiated, and the second derivative of the absorbance spectrum of the M kinds of raw materials obtained in step S 0 is used. Equation (1) is assumed to be the sum, and each coefficient W _k (k = 1 to M) is calculated by multiple regression analysis (step S2). Then, the value of each coefficient W _k obtained regarded as the volume of each component material, to calculate the respective weight multiplied by the density of the constituent material thereto (Step S3). Further, the obtained weight is multiplied by the calorie burned per unit weight of each constituent material to calculate the burned calorie of the plastic mixture 0 '(step S4). Thereafter, steps S1 to S4 are repeated.

As described above, according to the combustion calorie measuring apparatus A1 and the combustion calorie measuring method using the same according to the present embodiment, the collected waste plastic is crushed and reduced to a plurality of types of plastics. Produces a waste plastic mixture mixed in unknown proportions,
Since the calorific value of the plastic mixture can be measured before input, it is not necessary to sort waste plastics by type before and after measurement, and there is no problem in the storage location of each separated plastic. Also does not occur. As described above, it is possible to rationalize the apparatus and the processing procedure as compared with the related art.

[0015]

Embodiments [1. Modified Example of the Embodiment] In the combustion calorie measuring device A1 and the combustion calorie measuring method using the same according to the above embodiment, as an example of a method for calculating the combustion calorie of a plastic mixture, a mixture is first determined by using a multiple regression method. Although the method of calculating the volume of each constituent raw material and calculating the calorie burned of the entire mixture based on the calculated volume was described above, for example, a calorie estimation function (regression curve) was determined in advance by a predetermined evaluation criterion, By substituting, for example, the second derivative of the absorbance spectrum of the plastic mixture into the above function, it is also possible to directly calculate the calories burned of the plastic mixture. Hereinafter, as specific examples, the principal component regression method (hereinafter, referred to as PCR method),
A method of measuring calories burned using the partial least squares regression method (hereinafter, referred to as PLS method) will be described.

(A. Method of Measuring Burned Calories by PCR Method) It is assumed that an objective variable Y (for example, calories) can be represented by a linear combination of several explanatory variables X _k (for example, spectral values at each wavelength) as in the following equation. , The weight of its linear combination W _k
Is called “linear multiple regression analysis”. Y = W ₁ · X ₁ + W ₂ · X ₂ + ... + W _d · X _d = ΣW _i · X _i (i = 1 to d: number of explanatory variables) (2) In this case, the above W _k is determined. Contains the number of explanatory variables d
More data sets (Y, X _k ) are required. For example, when calculating the calorie burned from the absorbance spectrum, if the absorbance spectrum is selected as the explanatory variable, the data points usually reach several hundred points (for example, in increments of 2 nm). Measurements are required and are not realistic. Therefore, the above explanatory variables
The "principal component regression (PCR) method" uses the score u obtained by the principal component analysis.

Here, there are various types of the principal component analysis.
Because it is a widely known technique in the field,
An overview will be briefly described with reference to FIG. Trial
Material spectrum λ_k(K: measurement wavelength range, all N points)
Coordinate value X in N-dimensional space_kThink. This X_k(= Λ_k)
, Consider the following linear combination (this u is
A). u = p₁・ X₁+ P_Two・ X_Two+ ... + p_N・ X_N = Σ (p_k・ X_k) (3) Also, weight p_kLet t be a vector having as components. the above
The score u is a point (X₁, X_Two, ..., X_k)
(Shown by ○ in the graph on the right side of FIG. 8)
It is a value on the t-axis of the perpendicular line. When there are M samples
In this case, M points are plotted in the N-dimensional space, and the score u
Also M (u ₁~ U_M)Desired. Here, principal component analysis
Is a vector t that maximizes the following equation (this is the first principal component
Vector t1). S = Σu_j ^Two (I = 1 to M) (4) Subsequently, the vector orthogonal to the first principal component vector
That is, a vector that maximizes the above S is obtained, and this is
Let it be a component vector t2. Hereinafter, the Nth principal component vector t
It is possible to find up to N.

When the regression equation is expressed using the score u obtained by the principal component analysis as described above, the number of samples is significantly smaller than when the regression equation (2) using the original variable X _k is used. Can determine the regression equation.
The regression equation using the score u by the principal component analysis described above as the explanatory variable is as follows. _{Y = W 1 · u 1 +} W 2 · u 2 + ... + W d · u d = ΣW i · X i (i = 1~d: Number of explanatory variables) ... (5) Here, adopted is described variables (mainly The number d of component scores) is optimized by several indexes for evaluating excess / deficiency (overfitting / underfitting).

Next, a specific procedure for calculating the calories burned using the PCR method will be described with reference to the flowchart shown in FIG. First, before the measurement of the plastic mixture is started, an absorbance spectrum (second derivative) of the reference sample (M type) is collected (step S10).
a). Then, a principal component analysis is performed using the absorbance spectrum of the reference sample, and principal component vectors t _{1 to} t _d are calculated (step S10b). Further, with respect to the burned calories Y _i of the reference sample and the score u _ik determined by the principal component vector, W _k that minimizes the following equation is obtained, and the regression equation (5) is determined (step S10c). . _{S = Σ {Y i -Σ (} u ik · W k)} 2 (i = 1~M, k = 1~d) ... (6) When the measurement of the plastic mixture is initiated, the spectroscope 12
The plastic mixture 0 'flowing on the conveyor 4
Is measured and sent to the calorie calculator 13 (step S11). In calorie calculator 13, the absorbance spectra of obtained from the spectrometer 12 the plastic mixture 0 'while second derivative to calculate the score u ₁ ~u _d for principal component vector obtained in step S10b (Step S12) . And
The resulting score u ₁ ~u _d are substituted into regression equation (5) to calculate the calories burned plastic mixture 0 '(step S13). Thereafter, steps S11 to S13 are repeated.

(B. Method of Measuring Burned Calories by PLS Method)
Method) The PLS method is, for example, “chemometrics” (Yoshi Miyashita)
Katsu and Shinichi Sasaki, Kyoritsu Shuppan Co., Ltd.) 55-72,
"Chemometrics" (Tetsuro Aijima, Maruzen Co., Ltd.) No. 1
It is already widely known as described on pages 16-118.
This is the way it was done. Therefore, the PLS method is used here.
Brief description of the burned calorie measurement method
You. In the PLS method, as in the PCR method described above,
The objective variable Y (eg, calories burned) is the explanatory variable t
_i(I = 1 to d) (Q:
Weighting factor). Y = Q₁・ T₁+ Q_Two・ T_Two+ ... + Q_d・ T_d= ΣQ_i・ T_i (I = 1 to d: number of explanatory variables) (7) In addition, as described below, the explanatory variable t_iIs the spectrum X_k(K
= 1 to N), the above PC
It is the same as the R method. In the CLS method, the explanatory variable t_iTo
Call it a latent variable. t_i= W_i1・ X₁+ W_i1・ X_Two+ ... + W_iN・ X_N= Σ (W_ik・ X_k(K = 1 to N: the number of spectrum points) (8) The difference between the PLS method and the PCR method is that_ikAnd Q_iDecision
It is a fixed method. That is, in the PCR method, the W_ikIs the main component
Is a vector, and all W_ikTo
First decided, then this W_ikWith one calculation
Q_iWhereas the PLS method determines the latent variable
While sequentially increasing the number d from 1 to W _ikAnd Q_iDecide
I will do it. That is, in Step 1, d = 1 and W_1kWhen
Q₁Is determined, and in step 2, d = 2 and step 1
W obtained by_1kAnd Q₁Using the value of_2kAnd Q_TwoDetermine
You. Hereinafter, by repeating this, all W_ikAnd Q
_iTo determine.

Next, a specific procedure relating to a calculation process of combustion calories using the PLS method will be described with reference to a flowchart shown in FIG. First, before the measurement of the plastic mixture is started, an absorbance spectrum (second derivative) of the reference sample (M type) is collected (step S20).
a). Subsequently, the spectrum values at N points obtained by the M kinds of reference samples are represented by an N × M matrix X. Similarly, the caloric values of the M kinds of reference samples are represented by a vector Y. The d sets of latent variables are represented by an N × d matrix T, and the loading coefficients are represented by a d × M matrix P. Furthermore, each represent the weighting factor Q _i vector q, E residuals X, the residuals Y at f. Here, the following PLS model is assumed. X = T · P + E (9a) Y = T · q + f (9b) Next, it is assumed that the number of latent variables is d = 1. In this case, the above equation (9b) is a simple regression equation. Here, the inner product S = Y · T, which is the covariance of Y and T, is considered, and W _1k is obtained by considering the case where S is the maximum as the optimal model. In addition, the latent variable is obtained by the above equation (8). P of the above equation (9a)
Can be obtained under the condition that the sum of squares of the residual E of X becomes minimum, and q in the above equation (9b) can be obtained under the condition that the sum of squares of the residual f of Y becomes minimum. continue,
Assuming that d = 2, the first term of T, P, and q in the above equations (9a) and (9b) is the value obtained at d = 1, and the second term is an unknown number. Do. Thereafter, the process is repeated for the number of explanatory variables d. By the above calculations, the above equations (7) and (8) are determined (step S20).
b to S20e). When the measurement of the plastic mixture is started, the spectroscope 12 measures the absorbance spectrum of the plastic mixture 0 'flowing on the conveyor 4 and sends it to the calorie calculator 13 (step S21). In the calorie calculator 13, the absorbance spectrum of the plastic mixture 0 'obtained from the spectrometer 12 is secondarily differentiated, and the value of each latent coefficient is calculated by the above equation (8) (step S22). Then, the obtained value of the latent coefficient is substituted into the above equation (7), and the combustion calorie of the plastic mixture 0 'is calculated (step S23). Thereafter, steps S21 to S23 are repeated.

[2. Other Modifications] (a. MSC Treatment) In the above example, the second derivative of the absorbance spectrum was used to remove the influence on the measurement result due to the difference in the particle size of the crushed material. Multiplic
ative Scatter Correction) processing can expect the same effect. When additive and multiplicative baseline fluctuations of the absorbance spectrum occur due to fluctuations in the particle size of the sample, etc., the true spectrum A (λ) and the measured spectrum B
The following expression is assumed as (λ), multiplication effect α, and addition effect β. B (λ) = α × A (λ) + β The MSC processing is a waveform processing method for estimating α and β and estimating A (λ) from B (λ). In particular,
The absorbance spectrum of a reference sample (for example, one having a uniform particle size) is defined as S (λ), α and β that minimize X in the following equation are estimated, and A (λ) is obtained. X = {S (λ) -A (λ)} ² = {[S (λ)-{B (λ) -β) / α}] ² FIGS. 3 shows a comparison between the values of the absorbance spectrum before and after the MSC treatment and the absorbance spectrum of the reference sample. It can be seen that the correction value of the measurement spectrum of the unknown sample by the MSC process almost coincides with the spectrum of the reference sample. Here, the results of estimating the calories burned using the above three regression methods for the various plastic mixtures pulverized into a powder form are shown in the case where the pre-processing (secondary differential processing, MSC processing) is not performed. MSC
FIG. 12 shows a case where the process is performed and a case where the second derivative process is performed (only the second derivative is used for the multiple regression method). In this case, since a powdery sample was used, the error due to the difference in particle size was small, and in each case, a good calorie estimation result was obtained. However,
In the case without the pre-processing, the variation of the estimated value (●) of the reference sample is larger than that in the case where the pre-processing has been performed, and the effectiveness of the second derivative processing or the MSC processing can be confirmed. It has been confirmed that when a pellet-like sample is used, the estimation error increases without pretreatment due to the variation in particle size, and effective calorie estimation cannot be performed.

(B. Correction Regarding Temporal and Spatial Spread of Sample) When the absorbance spectrum is measured by the spectroscope 12,
If the amount (spatial spread) of the sample existing in the field of view of the spectroscope 12 is always constant, the obtained absorbance spectrum can be evaluated on the same basis. On the other hand, the amount of the sample present in the field of view of the spectroscope 12 (spatial spread)
If the value increases or decreases, the amount of reflected light from the sample changes accordingly, and the absorbance also changes. Therefore, the obtained absorbance spectrum cannot be evaluated on the same basis as it is. Further, for example, when measuring while transporting a sample on a conveyor, it is necessary to consider variations in the amount (temporal spread) of the sample existing in the field of view of the spectroscope 12. Therefore, when the spatial spread or the temporal spread of the sample as described above is considered to fluctuate, the obtained absorbance spectrum is converted to a value in which the temporal and spatial spreads are made constant. It needs to be converted (normalized). The above-mentioned normalization can be performed based on, for example, “the amount of the sample that has passed through the field of view of the spectroscope 12”. Here, the “amount of the sample that has passed through the field of view of the spectroscope 12” may be measured by separately installing a camera or the like, but the “amount of waste plastic sent during the measurement time” may be used instead. It is possible. For example, the height of the obtained spectrum waveform may be increased or decreased according to the “amount of waste plastic sent at the measurement time”. However, the correction method based on the "amount of the sample passing through the field of view of the spectroscope 12" as described above may deteriorate accuracy if the number of samples is larger than necessary on the conveyor. There is. In this case, since the sample is always present in the entire field of view of the spectroscope 12, even if the speed of the conveyor changes, the amount of sample to be sent changes, but the obtained absorbance spectrum becomes almost the same. It is. Therefore, when performing the above-mentioned correction, the upper limit of the sample loading amount on the conveyor must be "the amount that can be regarded as being occupied by the sample without any gap in the visual field" (for example, experimentally obtained). There is. In addition, the above-mentioned second derivative and MSC
It has been confirmed that if processing is used as preprocessing, the above-described correction based on temporal and spatial spread can be performed simultaneously.

(C. Other Apparatus Configuration) The configuration of the measurement apparatus is not limited to the measurement apparatus A1 as shown in FIG. As described above, the mixture 0 'may be distributed from the mixer 3' onto the measurement cell 14 by a predetermined amount, and the measurement may be performed for each cell. The mixture 0 'after the measurement is introduced into the incinerator 6 through the blowing pipe 16 by, for example, a blower 15.

[0025]

As described above, the present invention relates to a method for measuring the calorie of combustion of a plastic mixture in which a plurality of types of plastic pieces are mixed. A mixture spectrum measuring step of measuring the absorbance spectrum of the plastic mixture, the absorbance spectrum of the plastic mixture obtained in the mixture spectrum measuring step, and the absorbance spectrum of a plurality of reference samples having known calories burned. A combustion calorie calculating method for calculating the calorie burned of the plastic mixture on the basis of the obtained evaluation function and the method for measuring the calorie burned of the plastic mixture. Waste plastic A waste plastic mixture in which multiple types of plastics are mixed at an unknown ratio by crushing and granulation is generated, and the calorie burned of the plastic mixture can be measured before input, and the type of waste plastic can be measured before and after the measurement. There is no need for processing such as separation for each case, and there is no problem with the storage location of each separated plastic. As described above, it is possible to rationalize the apparatus and the processing procedure as compared with the related art.

[Brief description of the drawings]

FIG. 1 is a schematic diagram showing a schematic configuration of a combustion calorie measurement A1 according to an embodiment of the present invention.

FIG. 2 is a flowchart showing an example of a procedure for calculating calorie burned when a multiple regression method is used.

FIG. 3 shows measurement results of absorbance spectra of PP, PET, and PS alone.

FIG. 4 shows the measurement results of an absorbance spectrum of a mixture of two kinds (PP: PET = 1: 1).

FIG. 5 shows a mixture of three types (PP: PET: PS = 1: 1:
Measurement result of absorbance spectrum of 1).

FIG. 6 shows measurement results of absorbance spectra of PE in three different states: plate, pellet, and powder.

FIG. 7 is an explanatory diagram of a correction process based on a secondary differentiation.

FIG. 8 is an explanatory diagram of principal component analysis.

FIG. 9 is a flowchart showing an example of a procedure for calculating calorie burned when the PCR method is used.

FIG. 10 is a flowchart showing an example of a combustion calorie calculation processing procedure when the PLS method is used.

FIG. 11 is a graph comparing the absorbance spectrum of an unknown sample before and after MSC treatment with the absorbance spectrum of a reference sample.

FIG. 12 shows the results of estimating the calories burned by using three regression methods for various plastic mixtures pulverized into a powder form.
FIG. 9 is a diagram showing a case in which processing is not performed, a case in which MSC processing is performed, and a case in which secondary differentiation processing is performed.

FIG. 13 is a calorie measurement A according to an embodiment of the present invention.
FIG.

[Explanation of symbols]

 0: Waste plastic 0 ': Plastic mixture 1: Crusher 2: Particle size sorter 3: Mixer 4, 5: Conveyor 6: Combustion furnace 11: Light source 12: Spectrometer 13: Calorie calculator

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Katsuya Takaoka 1-5-5 Takatsukadai, Nishi-ku, Kobe City, Hyogo Prefecture Inside Kobe Research Institute, Kobe Steel Ltd. Kobe Steel, Ltd.Kobe Research Institute, Kobe Steel Co., Ltd. 1 Kanazawacho, Kakogawa-shi Kobe Steel, Ltd. Kakogawa Works F-term (reference) 2G040 BA25 CA23 2G059 AA03 BB08 EE01 EE12 MM02

Claims (18)

[Claims] In a method for measuring the calorie of combustion of a plastic mixture in which a plurality of types of plastic pieces are mixed, an absorbance spectrum of the plastic mixture crushed or reduced to a predetermined size or less is measured. Based on the mixture spectrum measurement step, the absorbance spectrum of the plastic mixture obtained in the mixture spectrum measurement step, and an evaluation function predetermined based on the absorbance spectra of a plurality of reference samples having known calories burned, Calculating a combustion calorie of the plastic mixture by calculating a combustion calorie of the plastic mixture. 2. The method according to claim 1, further comprising a step of crushing or reducing the size of the plastic mixture to a predetermined size or less. 3. In the combustion calorie calculation step, the evaluation function approximates an absorbance spectrum of the plastic mixture with a linear sum of spectra obtained by weighting absorbance spectra of the plurality of kinds of plastics alone serving as the reference sample. The weight is calculated by a multiple regression method by applying the absorbance spectrum of the plastic mixture obtained in the mixture spectrum measuring step to the regression equation, and the plastic mixture is formed based on the weight. 2. The calorie ratio of each plastic is calculated, and the calorie burn of the plastic mixture is further calculated based on the volume ratio, the density of the plural kinds of plastics alone, and the calorie burnt of the plural kinds of plastics alone. Or the method for measuring the calorific value of a plastic mixture according to 2 above . 4. The combustion calorie calculation step, wherein the evaluation function is a regression equation obtained by a principal component regression method based on absorbance spectra of a plurality of plastic mixtures whose calorie as the reference sample is known. Item 3. The method for measuring the calorie burned of a plastic mixture according to Item 1 or 2. 5. The combustion calorie calculating step, wherein the evaluation function is a regression equation obtained by a partial least squares method based on absorbance spectra of a plurality of plastic mixtures whose calorie as the reference sample is known. Item 3. The method for measuring the calorie burned of a plastic mixture according to Item 1 or 2. 6. The method according to claim 1, further comprising the step of correcting the absorbance spectrum of said plastic mixture from the absorbance spectrum of said plastic mixture obtained in said mixture spectrum measuring step.
The method for measuring the calorific value of a plastic mixture according to any one of the above. 7. The method according to claim 6, wherein the processing in the absorbance spectrum correcting step is a second derivative operation of the absorbance spectrum. 8. The processing in the absorbance spectrum correcting step is performed so that the sum of squares of the difference between the absorbance spectrum obtained from the predetermined reference plastic and the absorbance spectrum of the plastic mixture obtained in the mixture spectrum measuring step is minimized. 7. The method for measuring the calorie burned of a plastic mixture according to claim 6, wherein the method is a process of correcting an absorbance spectrum of the plastic mixture. 9. The method according to claim 1, wherein the absorbance spectrum of the plastic mixture obtained in the mixture spectrum measuring step is corrected so that at least one of the spatial spread and the temporal spread of the plastic mixture is adjusted to a predetermined value. 9. The method for measuring the calorie burned of a plastic mixture according to any one of 8. 10. An apparatus for measuring the calorie of combustion of a plastic mixture in which a plurality of types of plastic pieces are mixed, wherein an absorbance spectrum of the plastic mixture crushed or reduced to a predetermined size or less is measured. Based on the mixture spectrum measuring means, the absorbance spectrum of the plastic mixture obtained by the mixture spectrum measuring means, and the evaluation function predetermined based on the absorbance spectra of a plurality of reference samples having known calories burned, A combustion calorie calculating device for calculating the calorie burned of the plastic mixture. 11. An apparatus for measuring calorie combustion of a plastic mixture according to claim 10, further comprising means for crushing or reducing the size of said plastic mixture to a predetermined size or less. 12. The combustion calorie calculating means,
The evaluation function is a regression equation obtained by approximating the absorbance spectrum of the plastic mixture with a linear sum of spectra obtained by weighting the absorbance spectra of the plurality of kinds of plastics alone as the reference sample. The obtained absorbance spectrum of the plastic mixture is applied to the regression equation to calculate the weight by a multiple regression method, and the volume ratio of each plastic constituting the plastic mixture is calculated based on the weight,
12. The calorie burned of the plastic mixture according to claim 10 or 11, wherein the calorie burned of the plastic mixture is calculated based on the volume ratio, the density of the plurality of kinds of plastics alone, and the calorie burned of the plurality of kinds of plastics alone. measuring device. 13. The combustion calorie calculating means,
12. The combustion calorie measurement of a plastic mixture according to claim 10 or 11, wherein the evaluation function is a regression equation obtained by a principal component regression method based on absorbance spectra of a plurality of plastic mixtures whose calorific value is known as the reference sample. apparatus. 14. The combustion calorie calculating means,
12. The combustion calorie measurement of a plastic mixture according to claim 10 or 11, wherein the evaluation function is a regression equation obtained by a partial least squares method based on the absorbance spectra of a plurality of plastic mixtures whose calorific value is known as the reference sample. apparatus. 15. From the absorbance spectrum of the plastic mixture obtained by the mixture spectrum measuring means,
11. An apparatus according to claim 10, further comprising an absorbance spectrum correcting means for removing the influence of the particle size distribution of the plastic mixture.
15. The combustion calorie measuring device for a plastic mixture according to any one of items 14 to 14. 16. The apparatus according to claim 15, wherein the processing in the absorbance spectrum correcting means is a second derivative operation of the absorbance spectrum. 17. The processing in the absorbance spectrum correcting means such that the sum of squares of the difference between the absorbance spectrum obtained by the predetermined reference plastic and the absorbance spectrum of the plastic mixture obtained by the mixture spectrum measuring means is minimized. 16. The apparatus for measuring the calorie burned of a plastic mixture according to claim 15, wherein the apparatus is a processing for correcting an absorbance spectrum of the plastic mixture. 18. The method according to claim 10, wherein the absorbance spectrum of the plastic mixture obtained by the mixture spectrum measuring means is corrected so that at least one of the spatial spread and the temporal spread of the plastic mixture is adjusted to a predetermined value. 18. The apparatus for measuring the calorie burned of a plastic mixture according to any one of 17.