CN114813463A - Method for predicting basic density of papermaking wood chips by near infrared spectrum without moisture interference - Google Patents
Method for predicting basic density of papermaking wood chips by near infrared spectrum without moisture interference Download PDFInfo
- Publication number
- CN114813463A CN114813463A CN202210609092.8A CN202210609092A CN114813463A CN 114813463 A CN114813463 A CN 114813463A CN 202210609092 A CN202210609092 A CN 202210609092A CN 114813463 A CN114813463 A CN 114813463A
- Authority
- CN
- China
- Prior art keywords
- moisture
- near infrared
- spectrum
- sample
- water
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N9/00—Investigating density or specific gravity of materials; Analysing materials by determining density or specific gravity
- G01N9/24—Investigating density or specific gravity of materials; Analysing materials by determining density or specific gravity by observing the transmission of wave or particle radiation through the material
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a method for predicting the basic density of a papermaking wood chip by near infrared spectrum without moisture interference. The method comprises the following steps: collecting a near infrared dynamic spectrum in the water loss process of a papermaking wood chip sample, and calculating a moisture correction factor through an external parameter orthogonalization algorithm (EPO); collecting a near infrared spectrum of the papermaking wood chips in a water saturation state, performing moisture correction on the water saturation spectrum through a moisture correction factor, and establishing a partial least squares regression (PLS) model between the moisture correction spectrum and the basic density; and (4) firstly carrying out moisture correction on the spectrum of the sample to be detected, and then inputting the corrected spectrum into the model to carry out basic density prediction. According to the method, a steady wood chip basic density prediction model is established by eliminating moisture interference information in the near infrared spectrum, so that the interference of sample moisture fluctuation on a prediction result is effectively reduced, and the adaptability and stability of the model in a complex application environment are improved.
Description
Technical Field
The invention relates to the technical field of spectrum detection, in particular to a method for predicting the basic density of a papermaking wood chip by using near infrared spectrum, which eliminates moisture interference.
Background
The basic density is an important performance index for evaluating the pulping performance of the paper-making wood chips. Under the condition of a certain volume of production equipment, the higher the basic density of wood chips is, the higher the productivity is, but the wood chips with the too high density are difficult to soften, the impregnation of the liquid medicine is not uniform, the unstable product quality is easily caused, and the production cost is increased. Therefore, in the pulping production process, the density of the raw material wood chips is necessary to be monitored, and the process parameters are reasonably adjusted according to the property change of the wood chips so as to ensure the production stability. However, the traditional wood basic density detection method has high requirements on experimental conditions and long test period, and cannot meet the requirement of rapid batch detection.
The near infrared spectrum has the advantages of high analysis speed, simple instrument, non-destructiveness, no need of pretreatment of a sample and the like, is a generally accepted rapid detection technology, has wide application prospect in the field of wood material property analysis, and can realize rapid detection of various indexes of chemical components, fiber forms and physical properties of wood by performing correlated modeling on the spectrum and the material property indexes of the sample by a chemometrics method. Because the near infrared spectrum is very sensitive to the characteristic absorption of the moisture of the sample, in order to avoid the interference of moisture change, the establishment of the near infrared model is usually carried out under relatively stable experimental conditions, and the state of the sample to be tested is required to be consistent with that of the modeling sample as much as possible. However, in the pulping and papermaking industry, the wood chip raw material source is complex, and the conditions of felling, transportation and storage are inconsistent, so that the fluctuation of the moisture content of the wood chip is large, and when the wood chip is subjected to near infrared analysis, a large amount of moisture interference information in a spectrum can seriously influence the prediction of a model on the basic density of a sample, and the application effect of the near infrared technology in actual production is limited.
Disclosure of Invention
In order to solve the interference of the moisture change of the wood chips on the near infrared analysis, the invention provides a method for predicting the basic density of the wood chips by using the near infrared spectrum, which eliminates moisture interference information in the spectrum based on an external parameter orthogonalization algorithm, and improves the applicability and the robustness of the model in a complex application environment while not reducing the prediction precision of the model. The method comprises the following steps:
(1) preparing k paper making wood chip samples, firstly drying the samples in an oven at 105 ℃ to constant weight, recording the oven dry mass (g) of the wood chips, then soaking the samples in water to saturation (namely a water saturation state), and measuring the water saturation volume (cm) of the wood chips in the water saturation state by adopting a drainage method 3 ) Taking the ratio of wood chip absolute dry mass to water saturation volume as wood chip basic density (kg.m) 3 ). Arranging samples from small to large according to the basic density value, and selecting j samples as a moisture correction set by adopting an interval sampling method for calculating a moisture correction factor; and the residual samples are used as a training set for establishing a basic density prediction model.
(2) Collecting near infrared spectrum of a sample with concentrated moisture correction in a water-saturated state, putting the wood chips into an oven, and slowly drying at 35 DEG CDrying and monitoring the moisture content of the wood chips, collecting a near infrared spectrum once when the moisture of the wood chips is reduced by 5%, raising the temperature of an oven to 105 ℃ when the moisture of the wood chips is not changed any more, completely drying the wood chips to enable the wood chips to be in an oven-dry state (namely, the moisture content is zero), and then collecting the near infrared spectrum of the wood chips in the oven-dry state; selecting another sample in the water correction set, and repeating the steps until the near infrared spectrum collection of all the samples in the water correction set under different water content conditions in the process from a water saturation state to an oven dry state is completed; finally, the spectra of all samples collected under different moisture conditions from water saturation to absolute dryness form a dynamic spectrum data set
The moisture correction factor is calculated using an External parameter orthogonalization algorithm (EPO).
(3) Collecting a near infrared spectrum of a training concentrated sample in a water saturation state, namely a water saturation spectrum; correcting the moisture saturation spectrum through a moisture correction factor to obtain a moisture correction spectrum; and then performing Partial least squares regression (PLS) modeling on the moisture correction spectrum and the corresponding basic density value of the training set sample, and establishing a moisture correction basic density prediction model.
(4) And (4) acquiring a near infrared spectrum of the sample to be detected, correcting the near infrared spectrum by using a moisture correction factor, and inputting the obtained moisture correction spectrum into the moisture correction basic density prediction model in the step (3) to predict the density of the sample to be detected.
In the step (1), the selected papermaking wood chip sample is common broad-leaved wood papermaking wood chips such as poplar wood chips, eucalyptus wood chips and the like, and the specifications of the wood chips are required to be that the length and the width are less than or equal to 30mm, and the thickness is less than or equal to 6 mm. The number j of samples in the moisture correction set accounts for about one sixth of the total number k of samples, and the number (k-j) of samples in the training set is not lower than 50.
In the step (2) and the step (3), the conditions and the method for collecting the near infrared spectrum of the wood chips are as follows: the spectral range is 1100-2300nm, the integral time is 80ms, the scanning times are 16 times, 10 measuring points are randomly selected on the surface of the wood chip sample to collect diffuse reflection signals, and the average value is taken as the original near infrared spectrum of the sample. And carrying out Savitzky-Golay convolution smoothing, standard normal transformation and first derivative processing on the original spectrum, and then carrying out subsequent calculation.
Further, the steps of calculating a moisture correction factor by adopting an external parameter orthogonalization algorithm and performing spectrum moisture correction are as follows:
(a) calculating dynamic difference spectra
Dynamic spectra of samples under different moisture conditions,
is the spectrum of the sample in the oven dry state.
(b) Singular value decomposition (USV) of covariance matrix of D T ＝svd(D T D) U (n × n) is a left singular matrix, S (n × m) is a singular value diagonal matrix, V (m × m) is a right singular matrix, m is the number of samples, n is the number of wavelength points,
(c) determining EPO factor number g, obtaining the first g column submatrix V (m × g) of V, and calculating the moisture interference information matrix
(d) The moisture correction factor P is calculated as I-Q, I being the identity matrix.
(e) Calculating the moisture correction spectrum X * X is the original spectrum, XP.
And (3) optimizing and determining the EPO factor number g and the PLS latent variable number Lvs by a leave-one-out cross-validation method when establishing the water correction basic density prediction model.
The step (4) adopts Root Mean Square Error (RMSE) and Determination coefficients (R) 2 ) As an evaluation index, the model prediction performance was evaluated.
Wherein m is the number of samples, y i Is a standard method of measuring the value of,
is a predicted value
Is the average of the measurements made by the standard method.
The invention has the beneficial effects that:
the water-saturated wood chips are closest to the state of the wood chips during the actual measurement of the basic density, so that the spectrum is collected under the water-saturated state of the wood chips to be more favorable for establishing a basic density prediction model, but useful information in the spectrum mainly comes from information loaded by the interaction of near infrared light and a wood chip fiber structure, and the prediction performance of the model can be influenced by a large amount of characteristic absorption of moisture in the spectrum as interference information.
For the influence of sample moisture on a near-infrared model, the existing solution is mainly a global modeling or a hierarchical model, the global modeling is to jointly use spectra acquired under different moisture conditions for establishing the global model to realize the robustness of the model on moisture change, the hierarchical modeling is to classify and independently model samples with large moisture difference, and then a model close to the moisture of the sample to be detected is called for prediction. However, the modeling workload of the global model and the hierarchical model is huge, and the cost of sample preparation and model maintenance is high. The invention can directly eliminate the moisture interference information in the spectrum by dynamically monitoring the spectrum change of a small amount of samples under different moisture conditions and calculating the moisture correction factor.
Drawings
FIG. 1 shows the moisture content variation of a sample of a moisture calibration set during water loss;
FIG. 2 is a graph of raw near infrared spectra of a moisture calibration set of samples under different moisture conditions;
FIG. 3 is a spectrum after standard normal transformation and first derivative preprocessing;
FIG. 4 is a graph of model cross-validation root mean square error under different combinations of g and Lvs;
FIG. 5 is a water correction factor matrix chart;
FIG. 6 is a water content correction effect graph of a training set saturated water spectrum;
FIG. 7 test set spectra under different moisture conditions. Moisture conditions: 62.15 to 67.11 percent of M1; m2, 48.31-56.00%; 32.11 to 46.80 percent of M3; 31.50 to 42.22 percent of M4; 19.05 to 22.92 percent of M5; m6: 9.48-10.17%
FIG. 8 test set moisture corrected spectra under different moisture conditions. Moisture conditions: 62.15 to 67.11 percent of M1; m2, 48.31-56.00%; 32.11 to 46.80 percent of M3; 31.50-42.22% of M4; 19.05 to 22.92 percent of M5; 9.48 to 10.17 percent of M6;
FIG. 9 shows the test set prediction results of the saturated water spectrum basic density prediction model under different water conditions;
FIG. 10 moisture corrected base density prediction model predicts results for test sets under different moisture conditions.
Detailed Description
The invention is described in further detail below with reference to the figures and specific embodiments.
1. Sample preparation:
the wood sample used in this example was 107 fast-growing poplar, which was harvested from Shandong province, Jinan City, 9 years old, and had an average breast diameter of 35 cm. Poplar was peeled and processed into a sample of wood chips, which were 20mm (longitudinal) x 20mm (chord) x 5mm (radial) in size, for a total of 80 pieces. Drying all samples at 105 deg.C completely, recording wood sheet absolute dry mass (g), soaking the samples in water until saturation, and measuring wood sheet water saturation volume (cm) by drainage method 3 ) From wood chipsThe ratio of dry mass to saturated water volume was taken as the basis density (kg. m) of the wood chips 3 )。。
And randomly drawing 20 test sets from the samples for detecting the predicted performance of the model. And then arranging the rest 60 samples from small to large according to the basic density, selecting one sample every five samples, selecting 10 samples as a water correction set for calculating a water correction factor, and finally taking the rest 50 samples as a training set for establishing a basic density prediction model. The statistical information for each sample set is listed in table 1.
TABLE 1 sample set basis Density statistics
2. Spectral acquisition and data processing
And (4) moisture correction set: collecting the near infrared spectrum of a sample in a water-saturated state, then putting the wood chips into an oven, slowly drying at 35 ℃, monitoring the moisture content of the wood chips, collecting a near infrared spectrum signal once when the moisture of the wood chips is reduced by 5%, when the moisture of the wood chips is not changed at 35 ℃, completely drying the wood chips by raising the temperature of the oven to 105 ℃, and collecting an oven dry spectrum. And finally, collecting 14 dynamic spectrums under different moisture conditions such as water saturation and absolute dryness of each sample, forming the dynamic spectrums of all the moisture correction set samples into a data set, and calculating moisture correction factors by using an external parameter orthogonalization algorithm.
Training set: collecting near infrared spectrum of a sample in a water saturation state, establishing a water saturation spectrum basic density prediction model based on a partial least squares regression algorithm, then performing water correction on the water saturation spectrum, establishing a water correction basic density prediction model, determining EPO factor g and PLS latent variable number Lvs by a leave-one-out cross verification method, correcting Root Mean Square Error (RMSEC) by the model, and correcting a decision coefficient (R) of model correction 2 C ) Cross validation Root Mean Square Error (RMSECV), cross validation decision coefficient (R) 2 CV ) And evaluating the modeling effect.
And (3) test set: the water-saturated wood chips are subjected to constant temperature and humidity (23 ℃, 50 percent R)H) And (4) naturally airing, wherein the near infrared spectrum of the wood chips is collected at intervals of a certain time (0h, 5h, 7h, 10h, 15h and 20h) in the airing process, and the moisture content of the wood chips is recorded. Finally, spectra (M1 to M6) of the test set samples under 6 different moisture conditions of water saturation to air drying are collected together. Predicting the test set sample by using a saturated water spectrum basic density prediction model and a moisture correction basic density prediction model respectively to predict a Root Mean Square Error (RMSEP) and a model prediction decision coefficient (R) 2 P ) And evaluating the prediction effect of the model.
3. Results and discussion
Fig. 1 and fig. 2 are the moisture content change of the moisture calibration set sample during the dehydration process and the acquired dynamic spectrum, respectively. It can be seen that the whole spectrum absorption signal of the wood chip has strong correlation with moisture change, the higher the moisture is, the stronger the absorption signal is, mainly because the moisture in the gaps of the wood chip increases the forward scattering of light, and the absorption rate of near infrared light is improved. In order to improve the interference of baseline drift and spectral peak overlap on spectral analysis, standard normal transformation and first derivative processing are carried out on an original spectrum, so that spectral characteristics are clearer. As shown in FIG. 3, the OH frequency doubling characteristic absorption band (1900-2100nm) of water molecules has a regular and significant change as the wood chip moisture is reduced.
Performing external parameter orthogonalization calculation on the dynamic spectrum data set to obtain a moisture correction factor P, and optimizing an EPO main factor number g and a PLS latent variable number Lvs based on a training set, wherein the optimization process is shown in FIG. 4, when g is less than 3, the prediction accuracy of a model established based on the moisture correction spectrum is not influenced, which shows that only moisture interference information in the spectrum is eliminated at the moment, so that g is finally determined to be 2, and Lvs is determined to be 3 as an optimal modeling parameter. Fig. 5 is a moisture correction factor P matrix diagram when g is 2, in which the region with higher weight of the correction factor is mainly concentrated on the moisture characteristic absorption band, which proves that the EPO algorithm can accurately acquire moisture interference information in the wood chip spectrum.
TABLE 2 modeling Effect of the basic Density prediction model
FIG. 6 is a diagram showing the effect of water correction on the training set water saturation spectrum, wherein the 1400nm and 1900nm bands are occupied by wide "water peaks", and after water correction, the water absorption is effectively eliminated, and at the same time, more characteristic absorption is shown in the 1800nm band of 1500-. To evaluate the effect of moisture correction on modeling, a basic density prediction model was built using the training set saturation spectra and moisture correction spectra, respectively, with the results shown in table 2. The corrected Root Mean Square Error (RMSEC) and the cross-validation Root Mean Square Error (RMSECV) of the saturated water spectrum model are respectively 11.06 kg.m 3 And 13.80kg m 3 And the method shows better prediction accuracy. This indicates that the information of the internal fiber structure of the wood chips carried by the spectra acquired in the water-saturated state has a strong correlation with the wood chip density. After the spectrum is subjected to moisture correction, the model prediction accuracy is not affected, which also shows that a large amount of moisture characteristic absorption in the saturated spectrum is redundant and useless for establishing a basic density prediction model.
The robustness of the model was externally verified using test set sample spectra collected under different moisture conditions, with the results shown in table 3. Due to the high sensitivity of the near infrared spectrum to moisture, the change of the wood chip moisture generates large spectral difference, so that the predictive performance of the moisture saturation spectral model is very unstable when predicting test set samples under different moisture conditions (fig. 7). Predicted Root Mean Square Error (RMSEP) for the test set samples in a saturated state was 12.42kg m 3 However, as the sample moisture content gradually decreased, a larger deviation was predicted (fig. 8). As shown in FIG. 9 and FIG. 10, after the moisture absorption information in the spectrum is eliminated by the moisture correction factor, the model robustness is significantly improved, and the predicted root mean square error of the test set samples under different moisture conditions is stabilized at 12.62-13.59kg · m 3 And the anti-interference capability to the moisture change of the wood chips is strong. The method corrects the influence of wood chip moisture change on the near infrared spectrum based on the EPO algorithm, establishes a stable basic density prediction model, improves the applicability of the model to a complex application environment, and is favorable for promoting the near infrared analysis technology in industrial productionAnd (5) application and popularization.
TABLE 3 prediction Effect of the basic Density prediction model on test set samples
It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Claims (8)
1. A method for predicting the basic density of a papermaking wood chip by near infrared spectrum without moisture interference is characterized by comprising the following steps: the method comprises the following steps:
(1) arranging k papermaking wood chip samples from small to large according to the basic density value, and selecting j samples as a moisture correction set by adopting an interval sampling method for calculating a moisture correction factor; the residual samples are used as a training set for establishing a basic density prediction model;
(2) collecting a near infrared spectrum of a sample with concentrated moisture correction in a water saturation state, heating the sample to gradually reduce the moisture content of the sample, and respectively collecting the near infrared spectrum of the sample once under the condition of different moisture contents; when the moisture content is not changed any more, completely drying the sample to an oven-dry state, and then collecting the near infrared spectrum of the sample in the oven-dry state; selecting another sample in the water correction set, and repeating the steps until the near infrared spectrum collection of all the samples in the water correction set under different water content conditions in the process from a water saturation state to an oven dry state is completed; forming all the collected near infrared spectra into a dynamic spectrum data set, and calculating a moisture correction factor by adopting an external parameter orthogonalization algorithm;
(3) acquiring a near infrared spectrum of a training concentrated sample in a water saturation state to obtain a water saturation spectrum; correcting the moisture saturation spectrum through a moisture correction factor to obtain a moisture correction spectrum; then performing partial least squares regression modeling on the moisture correction spectrum and the corresponding basic density value of the sample in the training set, and establishing a moisture correction basic density prediction model;
(4) and (4) acquiring a near infrared spectrum of the sample to be detected, correcting the near infrared spectrum by using a moisture correction factor, and inputting the obtained moisture correction spectrum into the moisture correction basic density prediction model in the step (3) to predict the density of the sample to be detected.
2. The method for predicting basis density of papermaking wood chips using near infrared spectroscopy for eliminating moisture interference according to claim 1, wherein: in the step (1), the basic density value of the papermaking wood chip sample is determined by the following method: drying the sample to an absolute dry state, then placing the sample in water to soak until the sample is saturated by water, and measuring the saturated water volume (cm) of the wood chips by adopting a drainage method 3 ) The wood chip basic density is taken as the ratio of wood chip absolute dry mass to water saturated volume.
3. The method for predicting basis density of paper making chips by near infrared spectroscopy eliminating moisture interference according to claim 1, wherein: in the step (1), the papermaking wood chip sample is poplar wood chip or eucalyptus chip; the size requirements of the papermaking wood chip sample are as follows: the length and width are less than or equal to 30mm, and the thickness is less than or equal to 6 mm.
4. The method for predicting basis density of paper making chips by near infrared spectroscopy eliminating moisture interference according to claim 1, wherein: in the step (1), the ratio of j to k is 1: 6; the k-j is more than or equal to 50.
5. The method for predicting basis density of paper making chips by near infrared spectroscopy eliminating moisture interference according to claim 1, wherein: in the step (2), the near infrared spectrum is collected every time the water content of the sample in the water correction set is reduced by 5%.
6. The method for predicting basis density of paper making chips by near infrared spectroscopy eliminating moisture interference according to claim 1, wherein: the effective spectral range of the near infrared spectrum is 1100-2300 nm; the collected near infrared spectrum is the spectrum of the original near infrared spectrum after pretreatment.
7. The method for predicting basis density of paper making chips by near infrared spectroscopy eliminating moisture interference according to claim 6, wherein: the preprocessing method comprises the steps of carrying out standard normal transformation and first derivative processing on the original near infrared spectrum.
8. The method for predicting basis density of paper making chips by near infrared spectroscopy eliminating moisture interference according to any one of claims 1 to 7, wherein: using the root mean square error RMSE and the coefficient of determination R 2 As an evaluation index, the performance of the moisture correction basic density prediction model was evaluated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210609092.8A CN114813463A (en) | 2022-05-31 | 2022-05-31 | Method for predicting basic density of papermaking wood chips by near infrared spectrum without moisture interference |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210609092.8A CN114813463A (en) | 2022-05-31 | 2022-05-31 | Method for predicting basic density of papermaking wood chips by near infrared spectrum without moisture interference |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114813463A true CN114813463A (en) | 2022-07-29 |
Family
ID=82520146
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210609092.8A Pending CN114813463A (en) | 2022-05-31 | 2022-05-31 | Method for predicting basic density of papermaking wood chips by near infrared spectrum without moisture interference |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114813463A (en) |
-
2022
- 2022-05-31 CN CN202210609092.8A patent/CN114813463A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| He et al. | Prediction of hot-water-soluble extractive, pentosan and cellulose content of various wood species using FT-NIR spectroscopy | |
| CN103776777B (en) | Method for identifying ginsengs with different growth patterns by using near infrared spectrum technology and determining content of components in ginsengs | |
| CN105300923A (en) | Correction method of measuring point free temperature compensation model during process of online application of near infrared spectrum analyzer | |
| CN102506731B (en) | Method for detecting reconstituted tobacco thickness in papermaking process by utilizing near infrared spectrums | |
| CN106053383A (en) | Near-infrared online detection method for tobacco processing process | |
| CN114088661B (en) | Tobacco leaf baking process chemical composition online prediction method based on transfer learning and near infrared spectrum | |
| WO1995031710A1 (en) | Spectrophotometric method to measure quality and strength parameters in trees, lumber, timber, chips, saw dust, pulp and paper | |
| Pope | Near-infrared spectroscopy of wood products | |
| Sun et al. | Water content detection of potato leaves based on hyperspectral image | |
| Zhu et al. | Prediction of wood property in Chinese Fir based on visible/near-infrared spectroscopy and least square-support vector machine | |
| CN113484278A (en) | Tomato comprehensive quality nondestructive testing method based on spectrum and principal component analysis | |
| CN106706554A (en) | Method for rapidly and nondestructively determining content of straight-chain starch of corn single-ear grains | |
| CN114813463A (en) | Method for predicting basic density of papermaking wood chips by near infrared spectrum without moisture interference | |
| CN110596038A (en) | Method for rapidly determining starch content of sweet potatoes | |
| CN115824996A (en) | Tobacco conventional chemical component general model modeling method based on near infrared spectrum | |
| CN113049526B (en) | Corn seed moisture content determination method based on terahertz attenuated total reflection | |
| CN105181639A (en) | Method for quickly determining content of pentosan in broad-leaved wood based on near infrared spectrum technology | |
| CN107884360B (en) | Cigarette paper combustion improver detection method | |
| CN110646371A (en) | Method for measuring water content of tobacco essence perfume | |
| CN112861413A (en) | Biomass water content measurement and modeling method based on near infrared spectrum principal component and neural network | |
| CN112666122A (en) | Method for rapidly detecting glucose and moisture content after corn straw blasting pretreatment | |
| CN105352909A (en) | Method for rapid determination of pulping material extraction matter content on basis of near infrared spectroscopy technology | |
| CN111487219A (en) | Method for rapidly detecting content of bergamot pear lignin based on near infrared spectrum technology | |
| Baliza et al. | Evaluation of the spatial variation in moisture content inside wood pieces during drying by NIR spectroscopy | |
| CN113340845A (en) | Quantitative detection method for protein in soybean based on support vector regression algorithm and terahertz absorption spectrum |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |