CN111767790A - Chromatographic peak identification method based on convolutional neural network - Google Patents
Chromatographic peak identification method based on convolutional neural network Download PDFInfo
- Publication number
- CN111767790A CN111767790A CN202010411013.3A CN202010411013A CN111767790A CN 111767790 A CN111767790 A CN 111767790A CN 202010411013 A CN202010411013 A CN 202010411013A CN 111767790 A CN111767790 A CN 111767790A
- Authority
- CN
- China
- Prior art keywords
- peak
- point
- neural network
- convolutional neural
- derivative
- 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
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 42
- 238000011176 pooling Methods 0.000 claims abstract description 16
- 239000013598 vector Substances 0.000 claims abstract description 7
- 238000001914 filtration Methods 0.000 claims abstract 2
- 230000006870 function Effects 0.000 claims description 26
- 238000005070 sampling Methods 0.000 claims description 9
- 230000004913 activation Effects 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 6
- 210000002364 input neuron Anatomy 0.000 claims description 6
- 210000004205 output neuron Anatomy 0.000 claims description 6
- 230000000630 rising effect Effects 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 238000013459 approach Methods 0.000 claims description 3
- 230000001174 ascending effect Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims description 2
- 238000013507 mapping Methods 0.000 claims description 2
- 238000012549 training Methods 0.000 abstract description 3
- 230000005540 biological transmission Effects 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000012937 correction Methods 0.000 description 3
- 238000004451 qualitative analysis Methods 0.000 description 3
- 238000001228 spectrum Methods 0.000 description 3
- 238000004587 chromatography analysis Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 238000012271 agricultural production Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000002238 attenuated effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000012847 fine chemical Substances 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000003208 petroleum Substances 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
- G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N30/00—Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation
- G01N30/02—Column chromatography
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N30/00—Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation
- G01N30/02—Column chromatography
- G01N30/86—Signal analysis
- G01N30/8624—Detection of slopes or peaks; baseline correction
- G01N30/8631—Peaks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Abstract
The invention discloses a chromatographic peak identification method based on a convolutional neural network, and belongs to the field of chromatogram identification. The method comprises the steps of processing a wavelet coefficient of a signal containing noise by using a wavelet transform method, filtering noise in a chromatogram, and performing baseline calibration on the denoised chromatogram; obtaining an output value by using a convolutional neural network and interacting an input data set with a non-convolutional kernel in a convolutional layer, and extracting features in each small part of the picture; reducing the dimensionality of a feature vector output by the convolutional layer in the pooling layer, reducing the number of training parameters, reducing the over-fitting phenomenon, only retaining the most useful picture information and reducing the transmission of noise; finally, a classifier of the number of the required classes is generated by applying the full-connection layer, and peaks in the chromatogram are identified; and simultaneously, marking the characteristic points of the chromatogram by combining the first derivative and the second derivative. The chromatographic peak identification method provided by the invention has the advantage that the identification accuracy is obviously improved.
Description
Technical Field
The invention belongs to the field of chromatographic peak identification, and particularly relates to a chromatographic peak identification method based on a convolutional neural network.
Background
Chromatographic analysis is one of important instrument analysis means, has high separation efficiency, high analysis speed, high sensitivity and accurate qualitative and quantitative analysis result on complex multi-component mixtures, is increasingly widely applied to various fields of petroleum, fine chemical industry, medicine, biochemistry, electric power, white spirit, mines, environmental science and the like, and becomes an indispensable important separation and analysis tool for departments of industrial and agricultural production, scientific research, teaching and the like. Identification of chromatographic peaks is the basis and key to qualitative analysis and quantitative calculation in chromatographic analysis. The accuracy of chromatographic peak identification affects the determination of the name and concentration of the final component. Chromatographs use chemical or physical differences in the components of a mixture to gradually separate the components. During this process, the components may not be completely separated due to external factors and other disturbances. The expression on the chromatogram map can show completely separated peaks and irregular peaks such as overlapping peaks, front shoulder peaks, back shoulder peaks, tailing peaks, even negative peaks and the like. The occurrence of these irregular peaks increases the difficulty of peak identification. The premise of calculating the peak area is to determine the specific area of each peak, which is basically to determine the starting point, the peak and the end point of the peak.
With the further development of scientific research, chromatographic peak identification tends to be developed automatically and intelligently. In recent years, experts and scholars at home and abroad propose a plurality of chromatographic peak identification methods, and currently, a Fourier deconvolution method, a derivative method and the like are used more frequently. Although the fourier deconvolution method does not shift the peak position, which is convenient for qualitative analysis, the method is more suitable for identifying the almost completely overlapped chromatographic peaks, and the parameter value needs to be manually selected, which causes inconvenience for peak identification. The characteristic points are searched by using a first derivative method, and the start point, the top point and the end point of each overlapped peak can be found. However, the first derivative is not easy to detect the shoulder type overlapping peak of the peak intensity maximum point. The method for searching the characteristic points of the chromatographic peak by combining the first derivative and the second derivative is used for searching the characteristic points of the chromatographic peak, although the characteristic points of a single peak can be well identified, for the condition of multi-peak overlapping, particularly the condition that a shoulder peak appears, the identification accuracy rate is greatly reduced, the identification often cannot be accurately identified, the characteristic points such as the peak top point of the shoulder peak are obtained, in addition, the method which is not commonly used comprises a fuzzy matching method, an immune algorithm, a pattern identification method and the like, but the method does not meet the requirement of the identification accuracy rate.
Disclosure of Invention
Aiming at the technical problems in the prior art, the invention provides a chromatographic peak identification method based on a convolutional neural network, which is reasonable in design, overcomes the defects of the prior art and has a good effect.
In order to achieve the purpose, the invention adopts the following technical scheme:
a chromatographic peak identification method based on a convolutional neural network comprises the following steps:
step 1: preprocessing a test data set;
step 2: performing convolution operation on the preprocessed image by using a plurality of convolution kernels on the convolution layer by using a convolution neural network, and performing feature extraction and feature mapping;
and step 3: in a pooling layer, namely a down-sampling layer, performing down-sampling operation on the convolved image to obtain a feature map;
and 4, step 4: in the full connection layer, performing full connection operation on the pooled images, namely the feature maps;
and 5: identifying the test data set by using the trained convolutional neural network CNN;
step 6: calculating a first derivative and a second derivative of the identified image;
and 7: calculating the starting point and the end point of each peak according to the recognition result of the convolutional neural network CNN, the first derivative and the second derivative;
and 8: and outputting the result of peak identification.
Preferably, in step 1, the pretreatment mainly comprises: (1) the collected data is transferred from the memory to the database table; (2) performing baseline correction on the chromatogram; (3) and carrying out noise removal on the chromatogram.
Preferably, in step 2, the preprocessed image is convolved according to equation (1):
where the asterisk indicates the convolution operation, X is the input to the convolution layer, and H is the convolution kernel.
Preferably, in step 3, the convolved image is down-sampled according to formula (2):
wherein β is the weight coefficient of the pooling layer, down () is the down-sampling function,
represents a bias;
size of feature graph:
wherein, W is the image width, H is the image height, F is the convolution kernel width and height, and S is the step length of each movement.
Preferably, in step 4, the fully-connected layer expands a feature map obtained after undergoing a plurality of convolution and pooling operations into feature vectors, and then performs category judgment by taking the feature vectors as input of a classifier; assume a l +1 th fully-connected layer with an input of
Output is as
Wherein n and m represent the number of input neurons and the number of output neurons, respectively; the weight connecting the ith input neuron and the jth output neuron is recorded as
Performing full join operation on the pooled images according to formula (5):
wherein f () represents an activation function;
adopting a ReLU function as an activation function, wherein the expression is shown as formula (6):
when the input signal is less than 0, the output is 0, and when the input signal is greater than 0, the output is equal to the input.
Preferably, in step 7, the characteristic points of the peaks are judged according to the primitive function f (t), the first derivative g (t), the second derivative s (t) and the recognition result of the convolutional neural network CNN:
if only a single chromatographic peak exists in the chromatogram, the identification method is as follows:
(1) before the peak start: the curve is gentle, no obvious ascending and descending trend exists or the shaking intensity is small, and G (t) and S (t) both approach to 0;
(2) peak start point judgment conditions: g (t) > 0, S (t) > 0, and the peak starting point appears;
(3) left inflection point judgment condition: g (t) is the maximum value, and s (t) is 0, this point is the left inflection point;
(4) peak top judgment conditions: f (t) is the maximum value, g (t) is 0, s (t) is the minimum value, and this point is the peak top;
(5) right inflection point judgment condition: g (t) is the minimum value, and s (t) is 0, this point is the right-turn point;
(6) peak end point judgment conditions: g (t) 0, s (t) 0, the peak end point appears;
if a plurality of chromatographic peaks appear continuously, then according to the chromatographic peak top position identified by the convolutional neural network CNN, respectively searching the nearest points of g (t) ═ 0 and s (t) ═ 0 to both sides, and the point where no sign change appears in g (t) and s (t), namely the starting point/end point of the peak;
because the chromatographic peaks are Gaussian curves, namely the curves are symmetrical about two sides of the vertex, the characteristic point on the other side can be found out according to the nearest starting point/end point and the vertex;
if a plurality of chromatographic peaks appear consecutively:
(7) continuous peak-valley point judgment conditions: f (t) rising edge, G (t) rising state and sign change, S (t) is greater than 0;
the continuous peak valley point is the intersection point of the two chromatographic peaks.
The invention has the following beneficial technical effects:
the invention provides a chromatographic peak identification method based on a convolutional neural network, which effectively improves the accuracy of chromatographic peak identification, particularly identification of shoulder characteristic points.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Fig. 2 is a schematic diagram of a color spectrum after denoising by using a wavelet transform method.
Fig. 3 is a schematic diagram of baseline correction of a denoised chromatogram by using a wavelet transform method.
Fig. 4 is a schematic diagram of a test data set identified by using the trained convolutional neural network CNN.
Fig. 5 is a diagram illustrating the calculated first derivative.
FIG. 6 is a schematic diagram of the calculated second derivative.
Fig. 7 is a schematic diagram showing the sequential appearance of a plurality of chromatographic peaks.
Fig. 8 is a chromatogram peak identification result chart.
Detailed Description
The invention is described in further detail below with reference to the following figures and detailed description:
the method uses an N2000 chromatographic workstation to carry out noise removal and baseline calibration pretreatment on data, the experimental language is Python, model training of CNN is realized, and identification marking is carried out on the pretreated chromatogram.
A chromatographic peak identification method based on a convolutional neural network, the flow of which is shown in fig. 1, comprising the following steps:
step 1: preprocessing a data set, which mainly comprises the following steps: (1) the collected data is transferred from the memory to the database table; (2) and carrying out noise removal on the chromatogram.
Randomly selecting a chromatogram, and denoising the image by using a wavelet transform method, as shown in FIG. 2;
(3) performing baseline correction on the denoised chromatogram by using a wavelet transform method, as shown in FIG. 3;
step 2: extracting features and feature maps from the training data set using a plurality of convolution kernels;
for a two-dimensional convolution, the expression is:
the selected convolution kernels are:
the convolution kernel (a) is used for extracting a unimodal peak top in the chromatogram;
convolution kernels (b), (c) are used to extract the shoulder peaks in the chromatogram.
And step 3: in a pooling layer, namely a down-sampling layer, performing down-sampling operation on the convolved image to obtain a feature map;
and (3) performing downsampling operation on the convolved image according to the formula (2):
wherein β is the weight coefficient of the pooling layer, down () is the down-sampling function,
represents a bias;
the common pooling modes include random pooling, maximum pooling and average pooling. The invention adopts a maximum pooling method, and saves more image texture information by extracting the maximum value point in the window;
size of output image after pooling operation:
wherein, W is the image width, H is the image height, F is the convolution kernel width and height, and S is the step length of each movement.
And 4, step 4: in the full-connection layer, performing full-connection operation relation on the pooled images, namely the feature maps;
the full-connection layer expands a feature map obtained after a plurality of convolution and pooling operations into feature vectors, and then the feature vectors are used as the input of a classifier to carry out category judgment; assume a l +1 th fully-connected layer with an input of
Output is as
Wherein n and m represent the number of input neurons and the number of output neurons, respectively; the weight connecting the ith input neuron and the jth output neuron is recorded as
Performing full join operation on the pooled images according to formula (5):
wherein f () represents an activation function; common activation functions are Sigmoid, Tanh, ReLU, etc.
The invention adopts a ReLU function as an activation function, and the expression of the ReLU function is shown as the formula (6):
when the input signal is less than 0, the output is 0, and when the input signal is greater than 0, the output is equal to the input.
The expression of the ReLU function shows that when the input signal is less than 0, the output is 0, when the input signal is greater than 0, the output is equal to the input, and the ReLU function can keep the gradient not to be attenuated when the input is greater than or equal to 0, so that the problem of gradient disappearance is effectively solved, the gradient saturation phenomenon of the Sigmoid function and the Tanh function is effectively solved, and the convergence speed is obviously higher than that of the Sigmoid function and Tanh function; meanwhile, the Sigmoid function and the Tanh function need to be subjected to exponential calculation, the calculation complexity is high, the ReLU function can obtain an activation value only by one threshold value, and the calculation amount is obviously reduced.
And 5: identifying the test data set by using the trained convolutional neural network CNN; as shown in fig. 4;
step 6: calculating a first derivative and a second derivative of the identified image;
the calculated first derivative image, as shown in fig. 5;
second derivative images, as shown in FIG. 6;
and 7: calculating the starting point and the end point of each peak according to the recognition result of the convolutional neural network CNN, the first derivative and the second derivative; the trend rules of three curves of the monochromatic spectrum peak are summarized in table 1;
summary table of trend rules of three curves of monochromatic spectrum peak
Judging the characteristic points of the peaks according to the original function F (t), the first derivative G (t), the second derivative S (t) and the recognition result of the convolutional neural network CNN:
if only a single chromatographic peak exists in the chromatogram, the identification method is as follows:
(1) before the peak start: the curve is gentle, no obvious ascending and descending trend exists or the shaking intensity is small, and G (t) and S (t) both approach to 0;
(2) peak start point judgment conditions: g (t) > 0, S (t) > 0, and the peak starting point appears;
(3) left inflection point judgment condition: g (t) is the maximum value, and s (t) is 0, this point is the left inflection point;
(4) peak top judgment conditions: f (t) is the maximum value, g (t) is 0, s (t) is the minimum value, and this point is the peak top;
(5) right inflection point judgment condition: g (t) is the minimum value, and s (t) is 0, this point is the right-turn point;
(6) peak end point judgment conditions: g (t) 0, s (t) 0, the peak end point appears;
although the first-order and second-order derivative methods which are mainstream at present can better identify each characteristic point of a chromatographic peak aiming at the condition of a single peak, the method cannot well identify the top point of the chromatographic peak for continuous chromatographic peaks.
As shown in fig. 7, for the identification of the main peak in the case of the occurrence of the continuous chromatographic peaks, the peak p of the main peak can be identified by using the method of combining the first derivative and the second derivative, but for the identification of the shoulder peak, at the maximum value p' of f (t), although g (t) is 0, it is obvious that s (t) is not a minimum value at this point, so the identification result of this method has a certain error.
If a plurality of chromatographic peaks appear continuously, then according to the chromatographic peak top position identified by the convolutional neural network CNN, respectively searching the nearest points of g (t) ═ 0 and s (t) ═ 0 to both sides, and the point where no sign change appears in g (t) and s (t), namely the starting point/end point of the peak;
because the chromatographic peaks are Gaussian curves, namely the curves are symmetrical about two sides of the vertex, the characteristic point on the other side can be found out according to the nearest starting point/end point and the vertex;
as shown in fig. 8, if a plurality of chromatographic peaks appear consecutively:
(7) continuous peak-valley point judgment conditions: f (t) rising edge, G (t) rising state and sign change, S (t) is greater than 0;
the continuous peak valley point is the intersection point of the two chromatographic peaks.
Wherein, the main peak has the following characteristic points:
| a | b | c | a’ | a’ |
| peak origin | Left inflection point | Peak top point | Right inflection point | Peak end point |
Each characteristic point of the secondary peak:
| m | o | n | n’ |
| peak origin | Peak top point | Right inflection point | Peak end point |
Where the s points are consecutive peak to valley points.
And 8: and outputting the result of peak identification.
It is to be understood that the above description is not intended to limit the present invention, and the present invention is not limited to the above examples, and those skilled in the art may make modifications, alterations, additions or substitutions within the spirit and scope of the present invention.
Claims (6)
1. A chromatographic peak identification method based on a convolutional neural network is characterized by comprising the following steps: the method comprises the following steps:
step 1: preprocessing a test data set;
step 2: performing convolution operation on the preprocessed image by using a plurality of convolution kernels on the convolution layer by using a convolution neural network, and performing feature extraction and feature mapping;
and step 3: in a pooling layer, namely a down-sampling layer, performing down-sampling operation on the convolved image to obtain a feature map;
and 4, step 4: in the full connection layer, performing full connection operation on the pooled images, namely the feature maps;
and 5: identifying the test data set by using the trained convolutional neural network CNN;
step 6: calculating a first derivative and a second derivative of the identified image;
and 7: calculating the starting point and the end point of each peak according to the recognition result of the convolutional neural network CNN, the first derivative and the second derivative;
and 8: and outputting the result of peak identification.
2. The convolutional neural network-based chromatographic peak identification method according to claim 1, wherein: in step 1, the pretreatment mainly comprises: (1) the collected data is transferred from the memory to the database table; (2) filtering noise in the chromatogram; (3) and carrying out baseline calibration on the denoised chromatogram.
3. The convolutional neural network-based chromatographic peak identification method according to claim 1, wherein: in step 2, the preprocessed image is convolved according to equation (1):
where the asterisk indicates the convolution operation, X is the input to the convolution layer, and H is the convolution kernel.
4. The convolutional neural network-based chromatographic peak identification method according to claim 1, wherein: in step 3, the convolved image is down-sampled according to equation (2):
wherein β is the weight coefficient of the pooling layer, down () is the down-sampling function,
represents a bias;
size of feature graph:
wherein, W is the image width, H is the image height, F is the convolution kernel width and height, and S is the step length of each movement.
5. The convolutional neural network-based chromatographic peak identification method according to claim 1, wherein: in step 4, the full connection layer expands the feature map obtained after convolution and pooling into feature vectors, and then the feature vectors are processedThe quantity is used as the input of a classifier to carry out class judgment; assume a l +1 th fully-connected layer with an input of
Output is as
Wherein n and m represent the number of input neurons and the number of output neurons, respectively; the weight connecting the ith input neuron and the jth output neuron is recorded as
Performing full join operation on the pooled images according to formula (5):
wherein f () represents an activation function;
adopting a ReLU function as an activation function, wherein the expression is shown as formula (6):
when the input signal is less than 0, the output is 0, and when the input signal is greater than 0, the output is equal to the input.
6. The convolutional neural network-based chromatographic peak identification method according to claim 1, wherein: in step 7, the characteristic points of the peaks are judged according to the primitive functions f (t), the first derivative g (t), the second derivative s (t) and the recognition result of the convolutional neural network CNN:
if only a single chromatographic peak exists in the chromatogram, the identification method is as follows:
(1) before the peak start: the curve is gentle, no obvious ascending and descending trend exists or the shaking intensity is small, and G (t) and S (t) both approach to 0;
(2) peak start point judgment conditions: g (t) > 0, S (t) > 0, and the peak starting point appears;
(3) left inflection point judgment condition: g (t) is the maximum value, and s (t) is 0, this point is the left inflection point;
(4) peak top judgment conditions: f (t) is the maximum value, g (t) is 0, s (t) is the minimum value, and this point is the peak top;
(5) right inflection point judgment condition: g (t) is the minimum value, and s (t) is 0, this point is the right-turn point;
(6) peak end point judgment conditions: g (t) 0, s (t) 0, the peak end point appears;
if a plurality of chromatographic peaks appear continuously, then according to the chromatographic peak top position identified by the convolutional neural network CNN, respectively searching the nearest points of g (t) ═ 0 and s (t) ═ 0 to both sides, and the point where no sign change appears in g (t) and s (t), namely the starting point/end point of the peak;
because the chromatographic peaks are Gaussian curves, namely the curves are symmetrical about two sides of the vertex, the characteristic point on the other side can be found out according to the nearest starting point/end point and the vertex;
if a plurality of chromatographic peaks appear consecutively:
(7) continuous peak-valley point judgment conditions: f (t) rising edge, G (t) rising state and sign change, S (t) is greater than 0;
the continuous peak valley point is the intersection point of the two chromatographic peaks.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010411013.3A CN111767790B (en) | 2020-05-15 | 2020-05-15 | Chromatographic peak identification method based on convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010411013.3A CN111767790B (en) | 2020-05-15 | 2020-05-15 | Chromatographic peak identification method based on convolutional neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111767790A true CN111767790A (en) | 2020-10-13 |
| CN111767790B CN111767790B (en) | 2024-01-23 |
Family
ID=72720728
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010411013.3A Active CN111767790B (en) | 2020-05-15 | 2020-05-15 | Chromatographic peak identification method based on convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111767790B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112444589A (en) * | 2020-12-04 | 2021-03-05 | 深圳普门科技股份有限公司 | Chromatographic peak detection method, device, computer equipment and storage medium |
| CN113361436A (en) * | 2021-06-16 | 2021-09-07 | 中国农业大学 | Automatic signal identification method adopting first derivative and countermeasure network |
| CN113610817A (en) * | 2021-08-11 | 2021-11-05 | 贵州中烟工业有限责任公司 | Characteristic peak identification method, computing device and storage medium |
| CN115950993A (en) * | 2023-03-15 | 2023-04-11 | 福建德尔科技股份有限公司 | Method for testing fluorine content in fluorine-nitrogen mixed gas |
| CN116973563A (en) * | 2023-09-22 | 2023-10-31 | 宁波奥丞生物科技有限公司 | Immunofluorescence chromatography determination method and device based on quadrature phase-locked amplification |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0395481A2 (en) * | 1989-04-25 | 1990-10-31 | Spectra-Physics, Inc. | Method and apparatus for estimation of parameters describing chromatographic peaks |
| US5121443A (en) * | 1989-04-25 | 1992-06-09 | Spectra-Physics, Inc. | Neural net system for analyzing chromatographic peaks |
| CN102193900A (en) * | 2011-07-01 | 2011-09-21 | 国电南京自动化股份有限公司 | Peak recognition algorithm based on first-order derivative characteristic |
| CN105891397A (en) * | 2015-01-26 | 2016-08-24 | 大连达硕信息技术有限公司 | Comprehensive-two-dimensional-gas-chromatography peak detecting method |
| CN109214250A (en) * | 2017-07-05 | 2019-01-15 | 中南大学 | A kind of static gesture identification method based on multiple dimensioned convolutional neural networks |
| CN109507347A (en) * | 2017-09-14 | 2019-03-22 | 湖南中烟工业有限责任公司 | A kind of chromatographic peak selection method |
-
2020
- 2020-05-15 CN CN202010411013.3A patent/CN111767790B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0395481A2 (en) * | 1989-04-25 | 1990-10-31 | Spectra-Physics, Inc. | Method and apparatus for estimation of parameters describing chromatographic peaks |
| US5121443A (en) * | 1989-04-25 | 1992-06-09 | Spectra-Physics, Inc. | Neural net system for analyzing chromatographic peaks |
| CN102193900A (en) * | 2011-07-01 | 2011-09-21 | 国电南京自动化股份有限公司 | Peak recognition algorithm based on first-order derivative characteristic |
| CN105891397A (en) * | 2015-01-26 | 2016-08-24 | 大连达硕信息技术有限公司 | Comprehensive-two-dimensional-gas-chromatography peak detecting method |
| CN109214250A (en) * | 2017-07-05 | 2019-01-15 | 中南大学 | A kind of static gesture identification method based on multiple dimensioned convolutional neural networks |
| CN109507347A (en) * | 2017-09-14 | 2019-03-22 | 湖南中烟工业有限责任公司 | A kind of chromatographic peak selection method |
Non-Patent Citations (1)
| Title |
|---|
| 史晏廷 等: "基于神经网络综合分析的变压器油色谱在线监测系统", 电工电气, no. 6 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112444589A (en) * | 2020-12-04 | 2021-03-05 | 深圳普门科技股份有限公司 | Chromatographic peak detection method, device, computer equipment and storage medium |
| CN112444589B (en) * | 2020-12-04 | 2021-10-08 | 深圳普门科技股份有限公司 | Chromatographic peak detection method, device, computer equipment and storage medium |
| CN113361436A (en) * | 2021-06-16 | 2021-09-07 | 中国农业大学 | Automatic signal identification method adopting first derivative and countermeasure network |
| CN113610817A (en) * | 2021-08-11 | 2021-11-05 | 贵州中烟工业有限责任公司 | Characteristic peak identification method, computing device and storage medium |
| CN113610817B (en) * | 2021-08-11 | 2024-03-26 | 贵州中烟工业有限责任公司 | Characteristic peak identification method, computing device and storage medium |
| CN115950993A (en) * | 2023-03-15 | 2023-04-11 | 福建德尔科技股份有限公司 | Method for testing fluorine content in fluorine-nitrogen mixed gas |
| CN115950993B (en) * | 2023-03-15 | 2023-07-25 | 福建德尔科技股份有限公司 | Method for testing fluorine content in fluorine-nitrogen mixed gas |
| CN116973563A (en) * | 2023-09-22 | 2023-10-31 | 宁波奥丞生物科技有限公司 | Immunofluorescence chromatography determination method and device based on quadrature phase-locked amplification |
| CN116973563B (en) * | 2023-09-22 | 2023-12-19 | 宁波奥丞生物科技有限公司 | Immunofluorescence chromatography determination method and device based on quadrature phase-locked amplification |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111767790B (en) | 2024-01-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111767790A (en) | Chromatographic peak identification method based on convolutional neural network | |
| CN108615010B (en) | Facial expression recognition method based on parallel convolution neural network feature map fusion | |
| CN110532900B (en) | Facial expression recognition method based on U-Net and LS-CNN | |
| CN112070727B (en) | Metal surface defect detection method based on machine learning | |
| Zhang et al. | Multi-scale signed recurrence plot based time series classification using inception architectural networks | |
| CN107977661A (en) | The region of interest area detecting method decomposed based on full convolutional neural networks and low-rank sparse | |
| Gupta et al. | FPGA based real time human hand gesture recognition system | |
| CN112819748A (en) | Training method and device for strip steel surface defect recognition model | |
| CN110659374A (en) | Method for searching images by images based on neural network extraction of vehicle characteristic values and attributes | |
| Zhu et al. | Image identification of cashmere and wool fibers based on the improved Xception network | |
| CN114820471A (en) | Visual inspection method for surface defects of intelligent manufacturing microscopic structure | |
| Prasetya et al. | Indonesian food items labeling for tourism information using Convolution Neural Network | |
| CN106886754A (en) | Object identification method and system under a kind of three-dimensional scenic based on tri patch | |
| CN114140373A (en) | Switch defect detection method based on LabVIEW deep learning | |
| CN109934835A (en) | Profile testing method based on the adjacent connection of deeply network | |
| Badura et al. | Advanced scale-space, invariant, low detailed feature recognition from images-car brand recognition | |
| CN113221660A (en) | Cross-age face recognition method based on feature fusion | |
| Al-Tamimi | Combining convolutional neural networks and slantlet transform for an effective image retrieval scheme | |
| CN116934762A (en) | System and method for detecting surface defects of lithium battery pole piece | |
| CN111275732A (en) | Foreground object image segmentation method based on deep convolutional neural network | |
| Li et al. | A novel art gesture recognition model based on two channel region-based convolution neural network for explainable human-computer interaction understanding | |
| CN114565772A (en) | Set feature extraction method and device, electronic equipment and storage medium | |
| Desai et al. | Image retrieval using wavelet based shape features | |
| Backes et al. | Medical image retrieval based on complexity analysis | |
| Shipko et al. | Edge detection in hyperspectral images |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |