CN105259135A - Near-infrared measurement method applicable to real-time on-line measuring-point-free temperature compensation - Google Patents
Near-infrared measurement method applicable to real-time on-line measuring-point-free temperature compensation Download PDFInfo
- Publication number
- CN105259135A CN105259135A CN201510826933.0A CN201510826933A CN105259135A CN 105259135 A CN105259135 A CN 105259135A CN 201510826933 A CN201510826933 A CN 201510826933A CN 105259135 A CN105259135 A CN 105259135A
- Authority
- CN
- China
- Prior art keywords
- infrared
- temperature
- online
- temperature compensation
- real
- 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.)
- Granted
Links
Abstract
The invention relates to a near-infrared measurement method applicable to real-time on-line measuring-point-free temperature compensation. The method comprises the following steps: measuring physical property parameter standard data by utilizing a laboratory method; acquiring the near infrared spectrum of a same sample under different temperature levels; performing preprocessing and principal component analysis on the acquired spectral data, and performing non-separated latent variable modeling on the temperature by using partial least squares so as to acquire a physical property parameter measurement value at the present moment; and acquiring a new spectral data set online, constructing an on-line recursive algorithm, updating the physical property parameter measurement value, and finishing near-infrared on-line measurement capable of achieving a measuring-point-free temperature compensation function. According to the method disclosed by the invention, the temperature serving as a non-separated latent factor variable participates in the near-infrared modeling process, and the recursive algorithm is formed based on the to-be-measured property parameters, so that the property parameter measurement under different temperatures can be finished by virtue of adaptation of the model on the temperature, and direct temperature measurement information and correlation calculation are not needed; and therefore, the established model has the high universality. The on-line recursive algorithm is not sensitive on sample temperature change, and the measurement result has a relatively small integrated error.
Description
Technical field
The present invention relates to real-time online without measuring point temperature compensation near-infrared spectrometers measuring method, be applicable to physical parameter influenced by ambient temperature, as the on-line real-time measuremen of fluid viscosity, material density, constituent concentration, food quality, agricultural product composition, medicine active constituent content, oil product of gasoline quality etc.
Background technology
Near-infrared spectrum technique, because its analysis speed is fast, little to sample broke, without chemical contamination, be almost applicable to the feature such as all kinds of sample analysis, polycomponent hyperchannel Simultaneously test, becomes a bright spot in on-line analysis instrument.And existing most of near infrared detection method is all off-line measurement, can not provide more comprehensive, real-time sample message on the one hand for production and quality testing department, off-line measurement can not realize computing machine on-line monitoring and the object controlled in real time on the other hand.Therefore how to utilize near-infrared spectrometers measuring technique, realize online automatic analysis in real time and detect, have important meaning to the economic benefit and social benefit improving enterprise.
When near-infrared spectrometers real-time online is applied, measurement result can be subject to such environmental effects.Research shows, the change of temperature can produce the skew of vibrational spectrum, make the measurement result of near infrared spectrum under specified temp, be only applicable to the sample quality analysis at this temperature, and undesirable for the on-line analysis effect of sample quality, this shortcoming greatly limit the application of near-infrared spectrometers real-time online measuring technology.Therefore, research thermal adaptability is strong, precision is high, robustness is good real-time online measuring technology, and can become near infrared technology the key of effective application on site.
Summary of the invention
The method that the present invention proposes, during for on-line measurement, temperature variation has larger impact to near-infrared measuring, sets up the online recursive algorithm with temperature compensation mechanism.Object is when using near infrared online to measure, insensitive to the change of sample temperature, and measuring error is little.
The present invention for achieving the above object, adopts following technical scheme:
Step of the present invention is divided into three parts.Part I, the experimental design of modeling data, spectral collection and initial near infrared parameter models of physical are set up; Part II, the pre-service of near infrared spectrum and the foundation of calibration model; Part III, constructs online recursive algorithm, completes the near infrared online had without measuring point temperature compensation function and measures.
The experimental facilities of modeling data comprises, and the sample cell (2) that (1) can regulate sample temperature temperature meter (3) the near infrared spectrum collection instrument (4) of displays temperature change can not produce the optic probe of obviously impact on sample temperature.(5) the computer recording device be connected with near infrared spectrum collection instrument.
Experiment of the present invention and data collection step as follows:
Experimental procedure one: minimum and maximum temperature value under the online condition of confirmatory sample.Temperature range is divided into multiple level value.Each temperature levels is generally greater than thermometric instruments resolution 5 times, to reach effective discrimination.
Experimental procedure two: under a standard temperature of defined, primary standard data are obtained to all samples physical parameter.
Respectively spectroscopic data is collected to same sample under different temperatures level.Record corresponding sample temperature value simultaneously.This temperature value is used for the change of confirmatory sample temperature.In the present invention, temperature is as an implicit variable, and the accurate record of temperature value itself not necessarily.
Temperature as non-separation imply variable factors compensation model set up and correction algorithm as follows:
Step one: form target optical spectral data set with the spectrum under different temperatures level, to the target optical spectrum set pre-service that to carry out with physical property parameter mode to be measured be target.These pre-service comprise the superposition of one or more following algorithms: first order derivative, second derivative, maximum-minimum sandards, and basic bottom line corrects, scatter correction, constant bias correction, etc.The determination of Preprocessing Algorithm is herein different with the state of physical parameter to be measured and sample.Fig. 2 example is a kind of high molecular polymer original spectrum at different temperatures.Fig. 3 is the first order derivative pre-service local spectrum under different temperatures.
Step 2: do pivot analysis (PCA) to spectrum after the pre-service produced above, rejects statistics exceptional value, makes the pivot pattern of whole pre-processed spectrum data all within a statistical certainty.Fig. 4 example is a kind of high molecular polymer PCA mode chart.
Step 3: based on spectrum after above pre-service, using physical parameter to be measured in the original analysis value of a set point of temperature as predictive variable, after pre-service, spectrum wave number is as independent variable.Physical parameter calibration model is set up with partial least squares algorithm (PLS):
P＝D
1y
1+D
2y
2+…D
ny
n
Herein, P is the measured value of physical property variable at required standard temperature, D
i, i=1,2 ... n is regression coefficient, y
ibe after pre-service spectrum at wave number i=1,2 ... the numerical value at n place.
Step 4: the spectroscopic data collection that online acquisition is new, and obtain corresponding laboratory raw data simultaneously.Following method is utilized to form recurrence correction algorithm:
(1) calculate next step to measure: P
r(k+1)=P (k)+K [L (k-1)-P (k-1)]
(2) by current revised predicted value, P
rk () assignment gave the measured value P (k-1) in a upper moment, repeat above step, do recurrence assignment operation.
P herein
rk () is the current near infrared physical measurement modified value with temperature compensation, P (k-1) is the near infrared physical measurement value that previous step is not revised, L (k-1) is last computation actual physical parameter value used, and K is modifying factor or digital filter.
In above-mentioned steps four, modifying factor or digital filter can be more general Statistic analysis and Logic judgment, or their combination calculates.
In above-mentioned steps four, when each step calculates, physical parameter near infrared correction used can be regenerated by the spectroscopic data upgraded.Whole computational algorithm forms the form of recurrence.
The present invention participates in temperature near infrared modeling process as the implicit variable factors of non-separation, thus when using near-infrared measuring, can rely on the adaptability of model to temperature itself complete different temperatures under physical measurement, do not need direct temperature metrical information and correlation computations, make set up model have better versatility.The recursive algorithm invented has the preferably adaptability to sample temperature and the change of other measuring condition.
Accompanying drawing explanation
Fig. 1 is without measuring point temperature compensation experimental provision
A kind of high molecular polymer of Fig. 2 is at the original spectrum of different temperatures
First order derivative pre-service local spectrum under Fig. 3 different temperatures
A kind of high molecular polymer pivot analysis of Fig. 4 and pattern abnormity point
The near-infrared measuring model of a kind of high molecular polymer viscosity of Fig. 5
A kind of high molecular polymer modeling of Fig. 6 spectrum wave-number range used
The online recurrence result of implementation that Fig. 7 has without measuring point temperature compensation compares
Fig. 8 on-line measurement implementation step block diagram
Embodiment
Below for a kind of high molecular polymer viscosity measurement, specific implementation method is described.This example does not form and limits the scope of the inventive method.
Whole implementation step block diagram as shown in Figure 8.
Step 1: gather representative sample, ensure that the physical parameter to be measured of sample can cover the scope measured and require.Total number of samples is at 40-60.
Step 2: utilize the laboratory equipment shown in Fig. 1, gathers the near infrared spectrum of each sample respectively under 24 DEG C, 35 DEG C, 50 DEG C, 60 DEG C, 70 DEG C five different temperatures levels, records experiment condition as temperature etc. simultaneously.The original spectrum gathered is shown in Fig. 2.
Step 3: carry out different pre-service to spectrum and compare, to determine the last preprocess method be suitable for.In example, first order derivative process is carried out to macromolecule high viscosity sample.Treatment effect as shown in Figure 3.The effect of first order derivative process mainly eliminates the drift up and down of spectrum, remains the information that temperature variation affects spectral shape simultaneously.
Step 4: do pivot analysis (PCA) to spectrum after the pre-service produced above, rejects statistics exceptional value, makes the pivot pattern of whole pre-processed spectrum data all within a statistical certainty.As shown in Figure 4, in pivot mode chart, there are three abnormity point, in rejecting, modeling process should be used further to.
Step 5: based on spectrum after above pre-service, using physical parameter to be measured in the original analysis value of defined temperature as predictive variable, after pre-service, spectrum wave number is as independent variable.Physical parameter calibration model is set up with partial least squares algorithm (PLS):
P＝D
1y
1+D
2y
2+…D
ny
n
Herein, P is the measured value of physical property variable at required standard temperature, D
i, i=1,2 ... n is regression coefficient, y
ibe after pre-service spectrum at wave number i=1,2 ... the numerical value at n place.Fig. 5 is a kind of viscosity near-infrared measuring model of high molecular polymer.The correlativity of model predication value and measured value is 0.997, model accuracy R
2be 0.995.Fig. 6 modeling spectrum wave-number range used.Modeling wavelength band 7243-4497cm used
-1.
Step 6: the spectroscopic data collection that online acquisition 10 is new, and obtain corresponding laboratory raw data simultaneously.
Step 7: error E (k)=L (the k)-P (k) calculating 10 samples in the past, and form an error time sequence
E(k-1),E(k-2),…E(k-10)。
Step 8: the computing of low pass dynamic filter is done to above-mentioned error time sequence, obtains one-step prediction value, be designated as B.
Step 9: calculate viscosity correction measured value: P
r=P+B
P is the current near infrared physical measurement value with temperature compensation herein.
Step 10: by current modified value P
rk () assignment gave the measured value P (k-1) in a upper moment, do recursive operation, repeat above step 6-9.
Fig. 7 is the results contrast example of algorithms of different.Sample temperature changes between 24-70 degree Celsius.Fixed temperature model is the physical measurement model of employing 50 degrees Celsius of establishment of spectrums.As in figure indicate, No. 64 sample sequence moment, bring into operation recursive algorithm of the present invention.Can find out that the change of fixed temperature model to temperature is responsive.The change of the inventive method to sample temperature has preferably insensitivity, and recursive calculation makes measurement result have less global error simultaneously.
Claims (8)
1. real-time online without a measuring point temperature compensation near-infrared spectrometers measuring method, it is characterized in that, comprise the following steps:
Step one: the physical parameter normal data obtaining multiple testing sample, and under different temperatures level, carry out the collection of near infrared spectrum;
Step 2: pre-service and statistics outlier processing are done to the near infrared spectrum gathered in step one;
Step 3: using physical property Experiment Parameter room to be measured normal data as predictive variable, with the spectrum wave number processed in step 2 for independent variable, set up geophysical parameter prediction model, obtains the measured value under current time set point of temperature;
Step 4: the spectroscopic data collection that online acquisition is new, utilizes online recursive algorithm to carry out correction to measured value and upgrades.
2. according to claim 1 be applicable to real-time online without measuring point temperature compensation near-infrared measuring method, it is characterized in that: different temperatures level described in step one is distinguished interval and is greater than temperature measuring set resolution 5 times.
3. according to claim 1 be applicable to real-time online without measuring point temperature compensation near-infrared measuring method, it is characterized in that: the Pretreated spectra described in step 2 comprises the superposition of one or more following algorithms: first order derivative, second derivative, maximum-minimum sandards, basis bottom line corrects, scatter correction, constant bias correction etc.
4. according to claim 1 be applicable to real-time online without measuring point temperature compensation near-infrared measuring method, it is characterized in that: the statistics outlier processing method described in step 2 is principle component analysis, make the pivot pattern of derivative spectrum data and pre-processed spectrum data all in statistical certainty.
5. according to claim 1 be applicable to real-time online without measuring point temperature compensation near-infrared measuring method, it is characterized in that: the model of geophysical parameter prediction described in step 3 set up adopt partial least squares algorithm carry out the linear regression that temperature is implicit variable.
6. according to claim 1 be applicable to real-time online without measuring point temperature compensation near-infrared measuring method, it is characterized in that: described in step 4, online recurrence correction algorithm is:
P
r(k+1)＝P(k)+K[L(k-1)-P(k-1)]
Wherein P
rk () is the measurement modified value that current time has temperature compensation, P (k-1) was the measured value in a upper moment, and L (k-1) is last computation actual physical property reference value used, and K is modifying factor or lower order filter.
7. online recurrence correction algorithm according to claim 6, is characterized in that: K can be more general Statistic analysis and Logic judgment or both combinations.
8. online recurrence correction algorithm according to claim 6, is characterized in that: described online recurrence correction algorithm is when each step calculates, and physical parameter calibration model used can be regenerated by the spectroscopic data upgraded.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510826933.0A CN105259135B (en) | 2015-11-24 | 2015-11-24 | Suitable for real-time online without measuring point temperature-compensating near-infrared measuring method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510826933.0A CN105259135B (en) | 2015-11-24 | 2015-11-24 | Suitable for real-time online without measuring point temperature-compensating near-infrared measuring method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105259135A true CN105259135A (en) | 2016-01-20 |
| CN105259135B CN105259135B (en) | 2018-02-23 |
Family
ID=55098905
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510826933.0A Active CN105259135B (en) | 2015-11-24 | 2015-11-24 | Suitable for real-time online without measuring point temperature-compensating near-infrared measuring method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105259135B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105300924A (en) * | 2015-11-24 | 2016-02-03 | 江南大学 | Measuring point-free compensation modeling method of near infrared calibration model under effect of temperature |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103426167A (en) * | 2013-07-09 | 2013-12-04 | 哈尔滨工程大学 | Hyperspectral real-time detection method based on recursive analysis |
| CN104732215A (en) * | 2015-03-25 | 2015-06-24 | 广西大学 | Remote-sensing image coastline extracting method based on information vector machine |
-
2015
- 2015-11-24 CN CN201510826933.0A patent/CN105259135B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103426167A (en) * | 2013-07-09 | 2013-12-04 | 哈尔滨工程大学 | Hyperspectral real-time detection method based on recursive analysis |
| CN104732215A (en) * | 2015-03-25 | 2015-06-24 | 广西大学 | Remote-sensing image coastline extracting method based on information vector machine |
Non-Patent Citations (5)
| Title |
|---|
| 20130531: "Assessing the temperature influence on the soluble solids content of watermelon juice as measured by visible and near-infrared spectroscopy and chemometrics", 《JOURNAL OF FOOD ENGINEERING》 * |
| KATHERINE A. BAKEEV: "《过程分析技术 针对化学和制药工业的光谱方法和实施策略》", 30 September 2014 * |
| LIJUAN XIE 等: "Prediction of titratable acidity, malic acid, and citric acid in bayberry fruit by near-infrared spectroscopy", 《FOOD RESEARCH INTERNATIONAL》 * |
| XUEDIAN ZHANG 等: "Influence and correction of temperature on optical measurement for fat and protein contents in a complex food model system", 《INFRARED PHYSICS & TECHNOLOGY》 * |
| 王加华 等: "苹果糖度近红外光谱分析模型的温度补偿", 《光谱学与光谱分析》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105300924A (en) * | 2015-11-24 | 2016-02-03 | 江南大学 | Measuring point-free compensation modeling method of near infrared calibration model under effect of temperature |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105259135B (en) | 2018-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105300923B (en) | Without measuring point model of temperature compensation modification method during a kind of near-infrared spectrometers application on site | |
| CN105486658B (en) | A kind of near-infrared physical parameter measurement method with no measuring point temperature compensation function | |
| CA2843157C (en) | Extrapolation of interpolated sensor data to increase sample throughput | |
| CN101929951A (en) | Method for distinguishing milk doped with ewe's milk by near infrared spectrum | |
| CN107958267B (en) | Oil product property prediction method based on spectral linear representation | |
| CN105334185A (en) | Spectrum projection discrimination-based near infrared model maintenance method | |
| CN105466885B (en) | Based on the near infrared online measuring method without measuring point temperature-compensating mechanism | |
| CN109324013A (en) | A method of it is quickly analyzed using Gaussian process regression model building oil property near-infrared | |
| CN106441632A (en) | Detection method and detection device for thermometer | |
| CN106560699A (en) | Combined detection method used for identification of producing area of Wuyi rock tea | |
| CN107703097A (en) | Utilize the method and its application of decay total reflection probe and the model of near infrared spectrometer structure fast prediction oil property | |
| CN104390932A (en) | Method for detecting moisture content of wood on basis of infrared differential spectrum technology | |
| CN109738380A (en) | A kind of high-spectrum remote-sensing judgment method of soil salinization degree | |
| CN104516991A (en) | Gamma sensor full-temperature range compensation method | |
| CN105259136B (en) | Near infrared correction without measuring point temperature correction | |
| CN105628646B (en) | A kind of online tar prediction of cigarette and method for early warning | |
| CA2978172C (en) | Detecting a transient error in a body fluid sample | |
| CN116985183B (en) | Quality monitoring and management method and system for near infrared spectrum analyzer | |
| CN114088661A (en) | Online prediction method for chemical components in tobacco leaf curing process based on transfer learning and near infrared spectrum | |
| CN105300924A (en) | Measuring point-free compensation modeling method of near infrared calibration model under effect of temperature | |
| CN111141809B (en) | Soil nutrient ion content detection method based on non-contact type conductivity signal | |
| CN105259135A (en) | Near-infrared measurement method applicable to real-time on-line measuring-point-free temperature compensation | |
| CN114441469A (en) | Calibration method and device of moisture meter and computer equipment | |
| CN109614714A (en) | A kind of method for establishing model of acidified milk fragrance aesthetic quality control | |
| CN113158558B (en) | High-speed railway roadbed continuous compaction analysis method, device and analyzer |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |