CN116992293B - Intelligent data processing method for chemiluminescent instrument - Google Patents
Intelligent data processing method for chemiluminescent instrument Download PDFInfo
- Publication number
- CN116992293B CN116992293B CN202311243336.6A CN202311243336A CN116992293B CN 116992293 B CN116992293 B CN 116992293B CN 202311243336 A CN202311243336 A CN 202311243336A CN 116992293 B CN116992293 B CN 116992293B
- Authority
- CN
- China
- Prior art keywords
- standard
- value
- wavelength data
- data value
- sequence
- 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.)
- Active
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 12
- 239000000243 solution Substances 0.000 claims abstract description 34
- 238000000034 method Methods 0.000 claims abstract description 28
- 238000013528 artificial neural network Methods 0.000 claims abstract description 23
- 239000012086 standard solution Substances 0.000 claims abstract description 23
- 238000012549 training Methods 0.000 claims abstract description 21
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 17
- 238000012545 processing Methods 0.000 claims abstract description 9
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000010606 normalization Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 2
- 230000008569 process Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000004020 luminiscence type Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000003908 quality control method Methods 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/75—Systems in which material is subjected to a chemical reaction, the progress or the result of the reaction being investigated
- G01N21/76—Chemiluminescence; Bioluminescence
-
- 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
- G06N3/08—Learning methods
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2201/00—Features of devices classified in G01N21/00
- G01N2201/12—Circuits of general importance; Signal processing
- G01N2201/129—Using chemometrical methods
- G01N2201/1296—Using chemometrical methods using neural networks
Abstract
The invention relates to the technical field of data processing, in particular to an intelligent data processing method for a chemiluminescent instrument, which comprises the following steps: the method comprises the steps of carrying out EMD decomposition on standard curves of standard solutions to obtain IMF component sequences corresponding to each standard curve, obtaining standard sequence values, obtaining extremum of each wavelength data value in the components, determining initial importance of the same wavelength data value by combining luminous signal intensities corresponding to the same wavelength data value in all components corresponding to each standard sequence value, obtaining weight corresponding to each standard sequence value, obtaining importance of the same wavelength data value, further determining splicing sequences corresponding to each standard curve, forming a training set, obtaining a neural network, and inputting splicing sequences corresponding to actual curves of solutions to be tested into the neural network to obtain concentration of the solutions to be tested. According to the invention, by selecting important data, the accuracy of the neural network in identifying the concentration of different solutions is greatly improved.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to an intelligent data processing method for a chemiluminescent instrument.
Background
A chemiluminescent instrument is an instrument for detecting and analyzing chemiluminescent phenomena. The photoelectric conversion device converts the optical signals generated by chemical reaction into electric signals, thereby realizing observation, quantitative analysis and quality control of the luminescence phenomenon. And the chemiluminescent instrument has wide application range and can be used for detection, analysis and research tasks in a plurality of fields such as life science, medicine, environmental monitoring, food safety and the like. The method has the advantages of rapidness, sensitivity, no damage and high-flux analysis, and provides a powerful tool for research and application in the related field.
The existing problems are as follows: the existing method for detecting the precision of the chemiluminescent instrument is to construct a standard curve by measuring a series of standard solutions with known concentrations, and obtain the solution concentration corresponding to the actual curve by comparing the actual curve with the standard curve. However, since the solution is the same substance, when the change of the luminous signal intensity of the solution under different concentrations is small, the recognition rate is low, which may lead to inaccurate detection results.
Disclosure of Invention
The invention provides an intelligent data processing method for a chemiluminescent instrument, which aims to solve the existing problems.
The invention discloses an intelligent data processing method for a chemiluminescent instrument, which adopts the following technical scheme:
one embodiment of the present invention provides an intelligent data processing method for a chemiluminescent instrument, the method comprising the steps of:
collecting luminous intensity signals of the solution to be detected and a plurality of standard solutions under different wavelength data values by using a chemiluminescent instrument to obtain an actual curve of the solution to be detected and a standard curve of the standard solutions; using EMD decomposition to obtain an IMF component sequence corresponding to each standard curve and standard sequence values in the IMF component sequences corresponding to all the standard curves;
determining extremum of each wavelength data value in each IMF component corresponding to the standard ordinal value according to the difference between luminous signal intensities corresponding to adjacent wavelength data values in each IMF component corresponding to the standard ordinal value;
determining the initial importance of the same wavelength data value in all IMF components corresponding to each standard sequence value according to the luminous signal intensity and extremum corresponding to the same wavelength data value in all IMF components corresponding to each standard sequence value;
determining the weight corresponding to each standard sequence value according to the difference between all IMF components corresponding to each standard sequence value; determining the importance of the same wavelength data value in all standard curves according to the weight corresponding to all standard ordinal values and the initial importance of the same wavelength data value in all IMF components corresponding to all standard ordinal values;
determining a splicing sequence corresponding to each standard curve according to the importance of the same wavelength data value in all the standard curves; forming a training set according to all the splicing sequences, and training through the training set to obtain a neural network; inputting a splicing sequence corresponding to the actual curve of the solution to be measured into the neural network to obtain the concentration of the solution to be measured.
Further, the step of using EMD decomposition to obtain an IMF component sequence corresponding to each standard curve and standard sequence values in the IMF component sequences corresponding to all standard curves includes the following specific steps:
EMD decomposition is carried out on each standard curve, and a plurality of IMF components decomposed by each standard curve are obtained;
sequencing all IMF components decomposed by each standard curve from high frequency to low frequency to obtain an IMF component sequence corresponding to each standard curve;
and in the IMF component sequences corresponding to all the standard curves, when the number of the same sequence values is equal to the number of the preset standard solutions, marking the sequence values as standard sequence values.
Further, determining the extremum of each wavelength data value in each IMF component corresponding to the standard ordinal value according to the difference between the intensities of the light emitting signals corresponding to the adjacent wavelength data values in each IMF component corresponding to the standard ordinal value, including the following specific steps:
recording any one wavelength data value in any one IMF component corresponding to any one standard sequence value as a target wavelength data value;
in the luminous signal intensities corresponding to all the adjacent wavelength data values of the target wavelength data value, if the luminous signal intensity corresponding to the target wavelength data value is neither a maximum value nor a minimum value, setting the extremum of the target wavelength data value as a preset extremum;
and if the luminous signal intensity corresponding to the target wavelength data value is the maximum value or the minimum value, determining the extremum of the target wavelength data value according to the difference between the luminous signal intensity corresponding to the target wavelength data value and the luminous signal intensity corresponding to the adjacent wavelength data value.
Further, if the emission signal intensity corresponding to the target wavelength data value is a maximum value or a minimum value, determining the extremum of the target wavelength data value according to the difference between the emission signal intensity corresponding to the target wavelength data value and the emission signal intensity corresponding to the adjacent wavelength data value, including the specific steps as follows:
if the luminous signal intensity corresponding to the target wavelength data value is the maximum value, subtracting the maximum value of the luminous signal intensities corresponding to all the adjacent wavelength data values of the target wavelength data value from the luminous signal intensity corresponding to the target wavelength data value, and marking the maximum difference corresponding to the target wavelength data value;
dividing the maximum difference corresponding to the target wavelength data value by the luminous signal intensity corresponding to the target wavelength data value, and recording the maximum difference as the extremum of the target wavelength data value;
if the luminous signal intensity corresponding to the target wavelength data value is a minimum value, subtracting the luminous signal intensity corresponding to the target wavelength data value from the minimum value of the luminous signal intensities corresponding to all the wavelength data values adjacent to the target wavelength data value, and marking the minimum difference corresponding to the target wavelength data value;
and dividing the minimum difference corresponding to the target wavelength data value by the minimum value in the luminous signal intensities corresponding to all the adjacent wavelength data values of the target wavelength data value, and marking the minimum value as the extremum of the target wavelength data value.
Further, determining the initial importance of the same wavelength data value in all IMF components corresponding to each standard sequence value according to the intensity and extremum of the light emitting signal corresponding to the same wavelength data value in all IMF components corresponding to each standard sequence value, including the following specific steps:
and counting the luminous signal intensity and extremum corresponding to the same wavelength data value in all IMF components corresponding to each standard ordinal value, and determining the initial importance of the same wavelength data value in all IMF components corresponding to each standard ordinal value according to the mean value of the extremum and the variance of the luminous signal intensity.
Further, the specific calculation formula corresponding to the initial importance of the same wavelength data value in all IMF components corresponding to each standard sequence value is determined according to the average value of the extremum and the variance of the luminous signal intensity, wherein the specific calculation formula is as follows:
wherein S is the initial importance of the same wavelength data value in all IMF components corresponding to the ith standard ordinal value, z is the number of IMF components corresponding to the ith standard ordinal value,for extremum in the jth IMF component corresponding to the ith standard ordinal value for the same wavelength data value, V is the variance of the luminous signal intensities in all IMF components corresponding to the ith standard ordinal value for the same wavelength data value.
Further, the determining the weight corresponding to each standard sequence value according to the difference between all IMF components corresponding to each standard sequence value comprises the following specific steps:
in all IMF components corresponding to each standard ordinal value, calculating cosine similarity of any two IMF components, and marking a difference value obtained by subtracting the cosine similarity of the two IMF components as the difference of the two IMF components;
and marking the average value of the differences of all IMF components corresponding to each standard ordinal value as the weight corresponding to each standard ordinal value.
Further, determining the importance of the same wavelength data value in all standard curves according to the weight corresponding to all standard ordinal values and the initial importance of the same wavelength data value in all IMF components corresponding to all standard ordinal values, including the following specific steps:
the product of the weight corresponding to each standard ordinal value and the initial importance of the same wavelength data value in all IMF components corresponding to each standard ordinal value is recorded as the importance of the same wavelength data value under each standard ordinal value;
and determining the importance of the same wavelength data value in all standard curves according to the importance of the same wavelength data value under all standard sequence values.
Further, according to the importance of the same wavelength data value under all standard sequence values, a specific calculation formula corresponding to the importance of the same wavelength data value in all standard curves is determined as follows:
where m is the importance of the same wavelength data value in all standard curves,for the weight corresponding to the xth standard ordinal value in the IMF component sequences corresponding to all standard curves, y is the number of standard ordinal values in the IMF component sequences corresponding to all standard curves,/the weight is the weight of the xth standard ordinal value>For the initial importance of the same wavelength data value in all IMF components corresponding to the xth standard ordinal value in the IMF component sequence corresponding to all standard curves, +.>For the importance of the same wavelength data value under the x-th standard sequence value in the IMF component sequences corresponding to all standard curves, norm () is a linear normalization function.
Further, determining a splicing sequence corresponding to each standard curve according to the importance of the same wavelength data value in all the standard curves, including the following specific steps:
recording a wavelength data value with importance larger than a preset importance threshold value as a characteristic wavelength data value;
performing curve fitting on the luminous signal intensities corresponding to all the characteristic wavelength data values in each standard curve by using a polynomial fitting method to obtain a fitted curve, and recording the fitted curve as a characteristic standard curve of each standard curve;
on a standard curve, sequentially counting the luminous signal intensity corresponding to each wavelength data value from left to right to obtain a standard sequence;
sequentially counting the luminous signal intensity corresponding to each wavelength data value from left to right on the characteristic standard curve to obtain a characteristic standard sequence;
and sequentially arranging the data in the characteristic standard sequence from left to right after the last data in the corresponding standard sequence to obtain a spliced sequence corresponding to the standard sequence.
The technical scheme of the invention has the beneficial effects that:
according to the embodiment of the invention, the standard curve of the standard solution is subjected to EMD decomposition to obtain an IMF component sequence corresponding to each standard curve, standard sequence values in the IMF component sequences corresponding to all standard curves are obtained, the extremum of each wavelength data value in each IMF component corresponding to the standard sequence values is determined according to the difference between the luminous signal intensities corresponding to adjacent wavelength data values in each IMF component corresponding to the standard sequence values, and then the initial importance of the same wavelength data value in all IMF components corresponding to each standard sequence value is determined by combining the luminous signal intensities corresponding to the same wavelength data value in all IMF components corresponding to each standard sequence value. According to the difference between all IMF components corresponding to each standard sequence value, determining the weight corresponding to each standard sequence value, so as to obtain the importance of the same wavelength data value in all standard curves, further determining the splicing sequence corresponding to each standard curve, forming a training set according to all the splicing sequences, training through the training set to obtain a required neural network, and inputting the splicing sequence corresponding to the actual curve of the solution to be tested into the required neural network to obtain the concentration of the solution to be tested. The method comprises the steps of obtaining component data under different scales through EMD decomposition, obtaining the variation degree of each wavelength along with the concentration variation under each scale through the component data under different scales, facilitating the subsequent extraction of characteristic wavelengths, weighting the wave band importance under each scale by taking the difference of all the component data under each scale as a weight, obtaining more accurate importance, facilitating the subsequent neural network identification, fitting the characteristic wavelengths to obtain a fitting curve, avoiding the influence of important characteristics caused by a small number of wavelengths on the characteristics in an original curve, and finally obtaining a splicing sequence by taking a characteristic standard curve and a standard curve formed by the important data as the neural network input, thereby greatly improving the concentration identification precision and accuracy of the neural network on different solutions.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of the steps of an intelligent data processing method for a chemiluminescent apparatus of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purposes, the following detailed description refers to specific embodiments, structures, features and effects of an intelligent data processing method for a chemiluminescent apparatus according to the present invention with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of an intelligent data processing method for a chemiluminescent instrument.
Referring to FIG. 1, a flowchart of steps in an intelligent data processing method for a chemiluminescent apparatus according to one embodiment of the present invention is shown, the method comprising the steps of:
step S001: collecting luminous intensity signals of the solution to be detected and a plurality of standard solutions under different wavelength data values by using a chemiluminescent instrument to obtain an actual curve of the solution to be detected and a standard curve of the standard solutions; and (3) using EMD decomposition to obtain an IMF component sequence corresponding to each standard curve and standard sequence values in the IMF component sequences corresponding to all the standard curves.
A standard solution with known class and concentration of z groups of solutions and a group of solutions to be tested with known current class and unknown concentration are obtained. And collecting luminous intensity signals of the solution to be detected and all standard solutions under different wavelength data values by using a chemiluminescent instrument, so that each group of solutions corresponds to a data curve, wherein the horizontal axis of the data curve is the wavelength data value, and the vertical axis of the data curve is the luminous signal intensity. The data curve corresponding to the standard solution is recorded as a standard curve. And recording a data curve corresponding to the solution to be measured as an actual curve. The number z of standard solutions set in this example is 5, and this is described as an example, and other values may be set in other embodiments, and this example is not limited thereto. It should be noted that, wavelength data values on any two standard curves in all standard curves are in one-to-one correspondence, and the wavelength data values in one-to-one correspondence are the same.
The known EMD decomposition is a method for decomposing an original curve into a plurality of IMF components under different scales, wherein each IMF component corresponds to data of one scale, namely one sampling frequency, if a certain wavelength can represent more information which changes along with concentration under different sampling frequencies, the wavelength is of great importance, the wavelength is taken as a characteristic wavelength, and the concentration recognition effect and accuracy which are easy to unknown concentration can be greatly improved by participating in the concentration recognition process.
Therefore, EMD decomposition is carried out on each standard curve, and a plurality of IMF components decomposed by each standard curve are obtained. The IMF components of the EMD decomposition are known to be arranged in order from high frequency to low frequency, thereby obtaining a sequence of IMF components corresponding to each standard curve.
Because the number of IMF components decomposed by EMD of each standard curve is different, when the number of the same sequence values is equal to the number z of the set standard solutions in the IMF component sequences corresponding to all the standard curves, the sequence values are marked as standard sequence values to obtain y standard sequence values with different sizes, and each standard sequence value has one IMF component in the IMF component sequence corresponding to each standard curve, namely each standard sequence value corresponds to z IMF components. Because the IMF components corresponding to the standard ordinal values include all standard curves, the present embodiment only analyzes the IMF components corresponding to the standard ordinal values.
Step S002: and determining the extremum of each wavelength data value in each IMF component corresponding to the standard ordinal value according to the difference between the luminous signal intensities corresponding to the adjacent wavelength data values in each IMF component corresponding to the standard ordinal value.
Because the scales of the z IMF components corresponding to each standard sequence value are the same in the IMF component sequences corresponding to all standard curves, the initial importance of each wavelength data value under each scale can be obtained through calculation of each IMF component.
Taking z IMF components corresponding to the ith standard ordinal value in the IMF component sequences corresponding to all the standard curves as an example, the process of calculating the initial importance of each wavelength data value is as follows:
it is to be noted that in the EMD decomposition, the number of data of each IMF component is the same as the number of raw data. And then marking any one wavelength data value in any one IMF component in the z IMF components corresponding to the ith standard sequence value as a target wavelength data value, comparing the luminous signal intensity corresponding to the target wavelength data value with the luminous signal intensity corresponding to the adjacent wavelength data value, if the luminous signal intensity corresponding to the target wavelength data value is an extremum, and the difference between the luminous signal intensity and other values is larger, the luminous signal intensity corresponding to the target wavelength data value is more prominent, and in the concentration identification process, the luminous signal intensity corresponding to the target wavelength data value is less likely to be confused with other luminous signal intensities, so that the identification error is caused, and the initial importance of the target wavelength data value is larger.
Specifically, in the light emission signal intensities corresponding to all the wavelength data values adjacent to the target wavelength data value, if the light emission signal intensity corresponding to the target wavelength data value is neither a maximum nor a minimum, the extremum d of the target wavelength data value is made equal to. Extremum>Equal to 0, as described by way of example, other values may be provided in other embodiments, and the present example is not limited.
In the light-emitting signal intensities corresponding to all the adjacent wavelength data values, if the light-emitting signal intensity corresponding to the target wavelength data value is the maximum value, the calculation formula of the extremum d of the target wavelength data value is:
where d is the extremum of the target wavelength data value, a is the luminous signal intensity corresponding to the target wavelength data value, and b is the maximum value of the luminous signal intensities corresponding to all the wavelength data values adjacent to the target wavelength data value.The minimum difference between the luminous signal intensities corresponding to the target wavelength data value and the adjacent wavelength data value is expressed as a great difference.Is normalized.
In the luminous signal intensities corresponding to all the adjacent wavelength data values, if the luminous signal intensity corresponding to the target wavelength data value is a minimum value, the calculation formula of the extremum d of the target wavelength data value is:
where d is the extremum of the target wavelength data value, a is the luminous signal intensity corresponding to the target wavelength data value, and c is the minimum value of the luminous signal intensities corresponding to all the wavelength data values adjacent to the target wavelength data value.The minimum difference between the emission signal intensities corresponding to the target wavelength data value and the adjacent wavelength data value is denoted as the minimum difference. />Is normalized.
According to the mode, the extremum of each wavelength data value in each IMF component corresponding to the ith standard sequence value is obtained.
Step S003: and determining the initial importance of the same wavelength data value in all IMF components corresponding to each standard sequence value according to the luminous signal intensity and extremum corresponding to the same wavelength data value in all IMF components corresponding to each standard sequence value.
Taking any one wavelength data value of z IMF components corresponding to the ith standard sequence value as an example, wherein the wavelength data value is q, and counting the z luminous signal intensities and z extremums corresponding to the wavelength data value. From this, it can be known that the calculation formula of the initial importance S of the wavelength data value in the z IMF components corresponding to the ith standard ordinal value is:
wherein S is the initial importance of the wavelength data value in the z IMF components corresponding to the ith standard ordinal value, z is the number of the set standard solutions, z is the number of the IMF components corresponding to the ith standard ordinal value,for the extremum of the wavelength data value corresponding to the jth IMF component corresponding to the ith standard ordinal value, V is the variance of the z luminous signal intensities corresponding to the wavelength data value in the z IMF components corresponding to the ith standard ordinal value.
What needs to be described is:the larger the value of the z extremum corresponding to the wavelength data value in the z IMF components corresponding to the ith standard ordinal value is, the larger the difference between the intensity of the luminous signals corresponding to the wavelength data value and the adjacent wavelength value in each IMF component is, that is, the more important the wavelength data value is. And the larger V is used for representing the larger difference of the luminous signal intensities corresponding to the same wavelength data value in all IMF components in the same scale in all standard solutions, namely the more important the wavelength data value is. Therefore use->The product of V represents the initial importance of the wavelength data value in the z IMF components corresponding to the ith standard ordinal value.
According to the mode, the initial importance of all the different wavelength data values in the z IMF components corresponding to the ith standard sequence value is obtained, and the initial importance of all the different wavelength data values in the z IMF components corresponding to the other standard sequence values is obtained.
Step S004: determining the weight corresponding to each standard sequence value according to the difference between all IMF components corresponding to each standard sequence value; and determining the importance of the same wavelength data value in all standard curves according to the weight corresponding to all standard ordinal values and the initial importance of the same wavelength data value in all IMF components corresponding to all standard ordinal values.
The larger the difference between the IMF components among the z IMF components corresponding to the same standard ordinal value, the larger the weight of the initial importance of any one wavelength data value should be, because the larger the difference between the IMF components among the z IMF components corresponding to the same standard ordinal value is, which indicates that the larger the IMF components corresponding to the same standard ordinal value can distinguish standard curves with different concentrations.
Taking z IMF components corresponding to the ith standard sequence value as an example, calculating cosine similarity of any two IMF components, wherein a method for calculating the cosine similarity is a known technology, and a specific method is not described herein. The range of the cosine similarity is [ -1,1], and the larger the cosine similarity is, the more similar the luminous signal intensity in any two IMF components is, so that the difference value of the cosine similarity of any two IMF components is subtracted and is recorded as the difference of any two IMF components.
According to the mode, the difference of all IMF components corresponding to the ith standard ordinal value is obtained, and the average value of the difference of all IMF components corresponding to the ith standard ordinal value is recorded as the weight corresponding to the ith standard ordinal value.
Taking a wavelength data value with a wavelength data value q as an example in all standard curves, the wavelength data value corresponds to a plurality of standard ordinal values, and the greater the initial importance of the wavelength data value under all IMF components corresponding to each standard ordinal value, and the greater the weight corresponding to the standard ordinal value, the greater the importance of the wavelength data value.
The calculation formula of the importance m of the wavelength data value in all standard curves is known as follows:
where m is the importance of the wavelength data value in all standard curves,for the weight corresponding to the xth standard ordinal value in the IMF component sequences corresponding to all standard curves, y is the number of standard ordinal values in the IMF component sequences corresponding to all standard curves,/the weight is the weight of the xth standard ordinal value>The first of the wavelength data value in the z IMF components corresponding to the xth standard ordinal value in the IMF component sequence corresponding to all standard curvesThe initial importance. Norm () is a linear normalization function that normalizes the data value to [0,1]Within the interval.
What needs to be described is: the greater the initial importance of the wavelength data value under all IMF components corresponding to each standard ordinal value, and the greater the weight corresponding to the standard ordinal value, the greater the importance of the wavelength data value, and thereforeThe significance of the same wavelength data value in IMF components of all standard curves under the same decomposition scale is recorded as the significance of the same wavelength data value under the xth standard order value, so that the normalized value of the sum of the significance of the same wavelength data value in IMF components of all standard curves under different decomposition scales is used>The significance of this wavelength data value in all standard curves.
In the manner described above, the importance of each wavelength data value in all standard curves is obtained.
Step S005: determining a splicing sequence corresponding to each standard curve according to the importance of the same wavelength data value in all the standard curves; forming a training set according to all the splicing sequences, and training through the training set to obtain a neural network; inputting a splicing sequence corresponding to the actual curve of the solution to be measured into the neural network to obtain the concentration of the solution to be measured.
The importance threshold value set in this example is 0.7, and this is described as an example, and other values may be set in other embodiments, and this example is not limited thereto. The wavelength data value having an importance greater than the importance threshold is noted as a characteristic wavelength data value.
And further counting the luminous signal intensities corresponding to all the characteristic wavelength data values in each standard curve, obtaining a fitting curve by a polynomial fitting method, and recording the fitting curve as a characteristic standard curve of each standard curve. It should be noted that, the standard curves are in one-to-one correspondence with the wavelength data values on the corresponding characteristic standard curves, and the one-to-one correspondence with the wavelength data values is the same. The polynomial fitting is a well-known technique, and a specific method is not described here.
And counting the luminous signal intensity corresponding to each wavelength data value in turn from left to right on the standard curve to obtain a standard sequence. And counting the luminous signal intensity corresponding to each wavelength data value on the characteristic standard curve from left to right in sequence to obtain a characteristic standard sequence. And sequentially arranging the data in the characteristic standard sequence from left to right in the last data in the corresponding standard sequence to obtain a spliced sequence. And forming a training set according to all the splicing sequences, and training through the training set to obtain the neural network. The neural network inputs each spliced sequence and outputs the corresponding concentration. It should be noted that, the concentration of each standard solution is known, and the standard curve of each standard solution corresponds to one splicing sequence, and the known concentration of each standard solution is used as the label of the corresponding splicing sequence, so that the training set contains the known concentrations of all standard solutions and the corresponding splicing sequences.
In this embodiment, a cyclic neural network is adopted, and a mean square error loss function is used in a training process, which is a known technique, and a specific method is not described here.
According to the mode, a splicing sequence corresponding to the actual curve of the solution to be detected is obtained, and the splicing sequence corresponding to the actual curve of the solution to be detected is input into the neural network to obtain the concentration of the solution to be detected. Therefore, important data are extracted through data analysis, and the concentration identification precision and accuracy of the neural network to different solutions are greatly improved.
The present invention has been completed.
To sum up, in the embodiment of the present invention, EMD decomposition is performed on the standard curve of the standard solution to obtain an IMF component sequence corresponding to each standard curve, and standard sequence values in the IMF component sequences corresponding to all standard curves are obtained, according to the difference between the intensities of the light-emitting signals corresponding to adjacent wavelength data values in each IMF component corresponding to the standard sequence values, the extremum of each wavelength data value in each IMF component corresponding to the standard sequence values is determined, and then, in combination with the intensities of the light-emitting signals corresponding to the same wavelength data value in all IMF components corresponding to each standard sequence value, the initial importance of the same wavelength data value in all IMF components corresponding to each standard sequence value is determined. According to the difference between all IMF components corresponding to each standard sequence value, determining the weight corresponding to each standard sequence value, so as to obtain the importance of the same wavelength data value in all standard curves, further determining the splicing sequence corresponding to each standard curve, forming a training set according to all the splicing sequences, training through the training set to obtain a required neural network, and inputting the splicing sequence corresponding to the actual curve of the solution to be tested into the required neural network to obtain the concentration of the solution to be tested. According to the invention, the characteristic standard curve and the standard curve formed by the important data are used for obtaining the spliced sequence, and the spliced sequence is used as the input of the neural network, so that the concentration identification precision and accuracy of the neural network to different solutions are greatly improved.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (8)
1. An intelligent data processing method for a chemiluminescent instrument, the method comprising the steps of:
collecting luminous intensity signals of the solution to be detected and a plurality of standard solutions under different wavelength data values by using a chemiluminescent instrument to obtain an actual curve of the solution to be detected and a standard curve of the standard solutions; using EMD decomposition to obtain an IMF component sequence corresponding to each standard curve and standard sequence values in the IMF component sequences corresponding to all the standard curves;
determining extremum of each wavelength data value in each IMF component corresponding to the standard ordinal value according to the difference between luminous signal intensities corresponding to adjacent wavelength data values in each IMF component corresponding to the standard ordinal value;
determining the initial importance of the same wavelength data value in all IMF components corresponding to each standard sequence value according to the luminous signal intensity and extremum corresponding to the same wavelength data value in all IMF components corresponding to each standard sequence value;
determining the weight corresponding to each standard sequence value according to the difference between all IMF components corresponding to each standard sequence value; determining the importance of the same wavelength data value in all standard curves according to the weight corresponding to all standard ordinal values and the initial importance of the same wavelength data value in all IMF components corresponding to all standard ordinal values;
determining a splicing sequence corresponding to each standard curve according to the importance of the same wavelength data value in all the standard curves; forming a training set according to all the splicing sequences, and training through the training set to obtain a neural network; inputting a splicing sequence corresponding to an actual curve of the solution to be measured into a neural network to obtain the concentration of the solution to be measured;
according to the difference between the luminous signal intensities corresponding to the adjacent wavelength data values in each IMF component corresponding to the standard ordinal value, the extremum of each wavelength data value in each IMF component corresponding to the standard ordinal value is determined, and the specific steps are as follows:
recording any one wavelength data value in any one IMF component corresponding to any one standard sequence value as a target wavelength data value;
in the luminous signal intensities corresponding to all the adjacent wavelength data values of the target wavelength data value, if the luminous signal intensity corresponding to the target wavelength data value is neither a maximum value nor a minimum value, setting the extremum of the target wavelength data value as a preset extremum;
if the luminous signal intensity corresponding to the target wavelength data value is the maximum value or the minimum value, determining the extremum of the target wavelength data value according to the difference between the luminous signal intensity corresponding to the target wavelength data value and the luminous signal intensity corresponding to the adjacent wavelength data value;
if the light emitting signal intensity corresponding to the target wavelength data value is a maximum value or a minimum value, determining the extremum of the target wavelength data value according to the difference between the light emitting signal intensity corresponding to the target wavelength data value and the light emitting signal intensity corresponding to the adjacent wavelength data value, including the specific steps as follows:
if the luminous signal intensity corresponding to the target wavelength data value is the maximum value, subtracting the maximum value of the luminous signal intensities corresponding to all the adjacent wavelength data values of the target wavelength data value from the luminous signal intensity corresponding to the target wavelength data value, and marking the maximum difference corresponding to the target wavelength data value;
dividing the maximum difference corresponding to the target wavelength data value by the luminous signal intensity corresponding to the target wavelength data value, and recording the maximum difference as the extremum of the target wavelength data value;
if the luminous signal intensity corresponding to the target wavelength data value is a minimum value, subtracting the luminous signal intensity corresponding to the target wavelength data value from the minimum value of the luminous signal intensities corresponding to all the wavelength data values adjacent to the target wavelength data value, and marking the minimum difference corresponding to the target wavelength data value;
and dividing the minimum difference corresponding to the target wavelength data value by the minimum value in the luminous signal intensities corresponding to all the adjacent wavelength data values of the target wavelength data value, and marking the minimum value as the extremum of the target wavelength data value.
2. The method for intelligent data processing of a chemiluminescent apparatus of claim 1 wherein the steps of using EMD decomposition to obtain an IMF component sequence corresponding to each standard curve and standard sequence values in the IMF component sequences corresponding to all standard curves comprise the steps of:
EMD decomposition is carried out on each standard curve, and a plurality of IMF components decomposed by each standard curve are obtained;
sequencing all IMF components decomposed by each standard curve from high frequency to low frequency to obtain an IMF component sequence corresponding to each standard curve;
and in the IMF component sequences corresponding to all the standard curves, when the number of the same sequence values is equal to the number of the preset standard solutions, marking the sequence values as standard sequence values.
3. The method for intelligent data processing of a chemiluminescent apparatus of claim 1 wherein the determining of the initial importance of the same wavelength data value in all IMF components corresponding to each standard sequence value based on the intensity and extremum of the luminescent signal corresponding to the same wavelength data value in all IMF components corresponding to each standard sequence value comprises the following steps:
and counting the luminous signal intensity and extremum corresponding to the same wavelength data value in all IMF components corresponding to each standard ordinal value, and determining the initial importance of the same wavelength data value in all IMF components corresponding to each standard ordinal value according to the mean value of the extremum and the variance of the luminous signal intensity.
4. The method for intelligent data processing of a chemiluminescent apparatus of claim 3 wherein the specific calculation formula corresponding to the initial importance of the same wavelength data value in all IMF components corresponding to each standard sequence value is determined according to the mean of the extremum and the variance of the luminescent signal intensity:
wherein S is the initial importance of the same wavelength data value in all IMF components corresponding to the ith standard ordinal value, z is the number of IMF components corresponding to the ith standard ordinal value,for extremum in the jth IMF component corresponding to the ith standard ordinal value for the same wavelength data value, V is the variance of the luminous signal intensities in all IMF components corresponding to the ith standard ordinal value for the same wavelength data value.
5. The method for intelligent data processing of a chemiluminescent apparatus of claim 1 wherein the determining the weight for each standard sequential value based on the differences between all IMF components for each standard sequential value comprises the following steps:
in all IMF components corresponding to each standard ordinal value, calculating cosine similarity of any two IMF components, and marking a difference value obtained by subtracting the cosine similarity of the two IMF components as the difference of the two IMF components;
and marking the average value of the differences of all IMF components corresponding to each standard ordinal value as the weight corresponding to each standard ordinal value.
6. The method for intelligent data processing of a chemiluminescent apparatus of claim 1 wherein the determining the importance of the same wavelength data value in all standard curves based on the weights corresponding to all standard ordinal values and the initial importance of the same wavelength data value in all IMF components corresponding to all standard ordinal values comprises the following steps:
the product of the weight corresponding to each standard ordinal value and the initial importance of the same wavelength data value in all IMF components corresponding to each standard ordinal value is recorded as the importance of the same wavelength data value under each standard ordinal value;
and determining the importance of the same wavelength data value in all standard curves according to the importance of the same wavelength data value under all standard sequence values.
7. The method for intelligent data processing for a chemiluminescent apparatus of claim 6 wherein the specific calculation formula corresponding to the importance of the same wavelength data value in all standard curves is determined according to the importance of the same wavelength data value in all standard sequence values:
where m is the importance of the same wavelength data value in all standard curves,IMF score for all standard curvesWeight corresponding to the xth standard ordinal value in the quantity sequence, y is the number of standard ordinal values in the IMF component sequence corresponding to all standard curves, +.>For the initial importance of the same wavelength data value in all IMF components corresponding to the xth standard ordinal value in the IMF component sequence corresponding to all standard curves, +.>For the importance of the same wavelength data value under the x-th standard sequence value in the IMF component sequences corresponding to all standard curves, norm () is a linear normalization function.
8. The method for intelligent data processing of chemiluminescent apparatus of claim 1 wherein determining a splice sequence for each standard curve based on the importance of the same wavelength data value in all standard curves comprises the following steps:
recording a wavelength data value with importance larger than a preset importance threshold value as a characteristic wavelength data value;
performing curve fitting on the luminous signal intensities corresponding to all the characteristic wavelength data values in each standard curve by using a polynomial fitting method to obtain a fitted curve, and recording the fitted curve as a characteristic standard curve of each standard curve;
on a standard curve, sequentially counting the luminous signal intensity corresponding to each wavelength data value from left to right to obtain a standard sequence;
sequentially counting the luminous signal intensity corresponding to each wavelength data value from left to right on the characteristic standard curve to obtain a characteristic standard sequence;
and sequentially arranging the data in the characteristic standard sequence from left to right after the last data in the corresponding standard sequence to obtain a spliced sequence corresponding to the standard sequence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311243336.6A CN116992293B (en) | 2023-09-26 | 2023-09-26 | Intelligent data processing method for chemiluminescent instrument |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311243336.6A CN116992293B (en) | 2023-09-26 | 2023-09-26 | Intelligent data processing method for chemiluminescent instrument |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116992293A CN116992293A (en) | 2023-11-03 |
| CN116992293B true CN116992293B (en) | 2023-12-08 |
Family
ID=88523438
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311243336.6A Active CN116992293B (en) | 2023-09-26 | 2023-09-26 | Intelligent data processing method for chemiluminescent instrument |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116992293B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117368141B (en) * | 2023-12-07 | 2024-03-01 | 国检测试控股集团湖南华科科技有限公司 | Perchlorate wastewater concentration intelligent detection method based on artificial intelligence |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101852871A (en) * | 2010-05-25 | 2010-10-06 | 南京信息工程大学 | Short-term climate forecasting method based on empirical mode decomposition and numerical value set forecasting |
| CN101916439A (en) * | 2010-07-27 | 2010-12-15 | 哈尔滨工业大学 | Space debris hyperspectral sequence detection method based on Hilbert-Huang transform |
| KR101402213B1 (en) * | 2013-12-02 | 2014-06-02 | 대우조선해양 주식회사 | Method for modeling well log data using empirical mode decomposition(emd) and diffusion equation, recording media |
| CN110991721A (en) * | 2019-11-26 | 2020-04-10 | 国网山东省电力公司电力科学研究院 | Short-term wind speed prediction method based on improved empirical mode decomposition and support vector machine |
| CN114548158A (en) * | 2022-01-28 | 2022-05-27 | 广东工业大学 | Data processing method for blood sugar prediction |
| CN116359189A (en) * | 2023-03-29 | 2023-06-30 | 北京化工大学 | Neurotransmitter mixture quantitative determination method based on artificial neural network |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9565040B2 (en) * | 2014-07-01 | 2017-02-07 | The University Of New Hampshire | Empirical mode decomposition for spectrum sensing in communication systems |
-
2023
- 2023-09-26 CN CN202311243336.6A patent/CN116992293B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101852871A (en) * | 2010-05-25 | 2010-10-06 | 南京信息工程大学 | Short-term climate forecasting method based on empirical mode decomposition and numerical value set forecasting |
| CN101916439A (en) * | 2010-07-27 | 2010-12-15 | 哈尔滨工业大学 | Space debris hyperspectral sequence detection method based on Hilbert-Huang transform |
| KR101402213B1 (en) * | 2013-12-02 | 2014-06-02 | 대우조선해양 주식회사 | Method for modeling well log data using empirical mode decomposition(emd) and diffusion equation, recording media |
| CN110991721A (en) * | 2019-11-26 | 2020-04-10 | 国网山东省电力公司电力科学研究院 | Short-term wind speed prediction method based on improved empirical mode decomposition and support vector machine |
| CN114548158A (en) * | 2022-01-28 | 2022-05-27 | 广东工业大学 | Data processing method for blood sugar prediction |
| CN116359189A (en) * | 2023-03-29 | 2023-06-30 | 北京化工大学 | Neurotransmitter mixture quantitative determination method based on artificial neural network |
Non-Patent Citations (2)
| Title |
|---|
| Empirical mode decomposition of local field potential data from optogenetic experiments;Sorinel A. Oprisan等;《Frontiers inComputationalNeuroscience》;第 1-16y页 * |
| 基于经验模态分解和长短期记忆神经网络的 变压器油中溶解气体浓度预测方法;刘云鹏等;《中国电机工程学报》;第3998-4007页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116992293A (en) | 2023-11-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116992293B (en) | Intelligent data processing method for chemiluminescent instrument | |
| CN109870421B (en) | Incremental wood tree species classification and identification method based on visible light/near infrared spectrum analysis | |
| CN113177919B (en) | Lithology classification and principal component element content detection method combining LIBS and deep learning | |
| CN116434226B (en) | Circulating tumor cell analyzer | |
| CN110441254A (en) | A kind of near-infrared frequency comb spectrometer of plastics for identification | |
| CN109460471A (en) | A method of the mode based on self study establishes kinds of fibers spectrum library | |
| CN114660040A (en) | Method, device, medium and equipment for identifying microbial unicell species | |
| CN109470686A (en) | A kind of laser induced breakdown spectroscopy classifying identification method based on semi-supervised learning | |
| CN106198481B (en) | Fake wine identification device and method based on LIF technology and naive Bayes classification | |
| CN102967557B (en) | Biological aerosol water-content determination based on near infrared spectroscopy and classification method thereof | |
| CN115169504B (en) | Equipment abnormity identification method in coal gas fine desulfurization process | |
| CN107067034B (en) | Method and system for rapidly identifying infrared spectrum data classification | |
| CN107462330B (en) | Color recognition method and system | |
| CN111105041A (en) | Machine learning method and device for intelligent data collision | |
| CN104350378B (en) | Method and apparatus for the performance of measure spectrum system | |
| CN113435115B (en) | Fluorescence spectrum characteristic wavelength screening method, fluorescence spectrum characteristic wavelength screening device, computer equipment and readable storage medium | |
| CN115049917A (en) | Neural network training method, object detection method and intelligent terminal device | |
| CN109145887B (en) | Threshold analysis method based on spectral latent variable confusion discrimination | |
| CN117556245B (en) | Method for detecting filtered impurities in tetramethylammonium hydroxide production | |
| CN117316277B (en) | Gene detection data processing method based on fluorescence spectrum | |
| CN116990336B (en) | Method and system for determining purity of hexafluoroethane | |
| CN113539353B (en) | Method and system for determining chlorophyll fluorescence characteristic parameters under dark adaptation | |
| CN117007825B (en) | Reagent automatic identification and positioning system and method for full-automatic coagulation tester | |
| CN117129632A (en) | Method for detecting plant disease and pest infection by utilizing electronic nose | |
| CN114548183A (en) | Micro-seismic signal identification method based on online multi-core integration |
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 |