Bidirectional step-by-step local optimization-based aliasing spectrum analysis algorithm

Abstract

The invention discloses an aliasing spectrum analysis algorithm based on bidirectional step-by-step local optimization. The invention belongs to the technical field of marine water ecological environment monitoring. During analysis, an unknown to-be-analyzed aliasing three-dimensional fluorescence spectrum is locally optimized step by step according to a forward analysis sequence, and an initial solution combination is obtained; then, on the basis of a forward analysis result, reverse analysis is carried out, local re-optimization is carried out step by step in sequence, and a reverse re-analysis solution combination is obtained; comparing the two groups of positive and negative local optimal solutions, and judging whether a stopping condition is met or not; if not, the algae analysis sequence is changed and then analysis is carried out again; if yes, circulation is stopped, and the optimal analysis result is directly output. According to the two-way step-by-step optimization analysis algorithm, on the basis of ensuring the reliability of a measurement result, the operand is greatly saved, and the spectrum analysis efficiency is remarkably improved.

Description

Technical Field

[0001] The invention belongs to the technical field of marine water ecological environment monitoring, and in particular relates to an aliasing spectrum analysis algorithm based on bidirectional stepwise local optimization. Background Art

[0002] Algae is a general term for a class of microscopic algae that live in the water in a planktonic manner. Algae are widely distributed, and most of them are distributed in water. They are the material and energy basis of aquatic ecosystems and are also indicator organisms of water quality. Their species composition, community structure and abundance changes directly affect water quality, ecosystem energy flow, material flow and biological resource changes. As the most important primary producer in oceans and inland waters, the photosynthetic productivity of algae accounts for about 45% of the world's total plant productivity, and it also plays a very important role in maintaining the normal operation of the entire earth's ecosystem.

[0003] The in vivo fluorescence method has the advantages of fast measurement speed, high sensitivity, and no need for sample pretreatment. It is currently the most promising method for rapid on-site measurement of phytoplankton concentration in large areas of water. The basic idea is to invert the chlorophyll a concentration of algae through the fluorescence intensity generated by stimulated radiation of chlorophyll a molecules in living algal cells. At the same time, the photosynthetic pigment composition of different types of living algal cells has interspecies differences, and therefore has its own characteristic fingerprint fluorescence spectrum. When there are multiple components in the sample, within the range of Lambert-Beer's law, the algae fluorescence spectrum satisfies linear additivity, so the quantitative problem of algae concentration can be transformed into a linear regression problem of the matrix. This is the principle basis for the rapid classification and quantitative measurement of phytoplankton through in vivo fluorescence spectroscopy.

[0004] However, as a byproduct of the photosynthesis process of phytoplankton, the yield of live fluorescence is closely related to the absorption, transmission and release efficiency of light energy in the photosynthetic process of living cells. The instability of live fluorescence of phytoplankton has an adverse effect on the accurate quantification of the live fluorescence chlorophyll a concentration of phytoplankton. By constructing an interpolation standard spectral library for phytoplankton, the classification accuracy and measurement accuracy of the multivariate regression analysis measurement results can be significantly improved. However, it also brings new problems: when the capacity of the iterative standard spectral library of phytoplankton of different classes is large, if the global traversal method is adopted, it is easy to cause an explosion of computational workload, and the spectral analysis algorithm is also difficult to quickly converge to the true value of the mixed sample. This makes it difficult to apply the interpolation spectral library construction method to the analysis of large batches of samples, affecting the application of online analysis fluorescence instruments for phytoplankton in large areas of water. Summary of the invention

[0005] In view of the problems in the background technology, the present invention proposes an aliasing spectrum analysis algorithm based on bidirectional step-by-step local optimization. During analysis, the present invention gradually performs local optimization on the unknown aliased three-dimensional fluorescence spectrum to be analyzed in the forward analysis order, and obtains the initial solution combination; then, based on the result of the forward analysis, the reverse analysis order is gradually performed to perform local re-optimization, and a reverse re-analysis solution combination is obtained; the two groups of forward and reverse local optimal solutions are compared to determine whether the termination condition is met; if not, the algae analysis order is changed and re-analyzed; if yes, the cycle is terminated and the optimal analysis result is directly output.

[0006] The technical solution of the present invention is as follows: An aliasing spectrum analysis algorithm based on bidirectional step-by-step local optimization includes the following steps: Step 1: Analyze the three-dimensional fluorescence spectrum using the non-negative least squares method. When analyzing, under the condition that the standard spectrum of phytoplankton from the second to the last phylum of phytoplankton is the average value of all spectra in the standard spectrum library, the standard spectrum library of the first phylum of phytoplankton is optimized in the entire library: , in: Represents the unknown phytoplankton in vivo three-dimensional fluorescence spectrum to be analyzed; represents the average of all spectra in the canonical spectral library of phytoplankton from the second to the last phylum of phytoplankton (Cryptophyta, Chlorophyta, Bacillariophyta, Dinophyta, and Xanthophyta); The linear regression analytical coefficients representing the chlorophyll a concentration of the second to last phylum of phytoplankton; It means to traverse the minimum residual standard spectrum in the standard spectrum library of phytoplankton of the first phylum (Cyanobacteria); represents the least square analytical coefficient of the chlorophyll a concentration of the first phylum, phytoplankton; represents the minimum residual.

[0007] Step 2: Analyze the three-dimensional fluorescence spectrum of the living body Deduct the fluorescence intensity of the first phylum (Cyanobacteria) from the equation to obtain the difference spectrum to be analyzed. ; Under the condition that the standard spectrum of the first phylum (cyanobacteria) of phytoplankton is the minimum residual standard spectrum, and the standard spectrum of the third to the last phylum of phytoplankton is the average value of all spectra in the standard spectrum library, the difference spectrum to be analyzed is analyzed by the non-negative least squares method: , , in: It represents the minimum residual normalized spectrum in the normalized spectrum library of phytoplankton of the second phylum (Cryptophyta); represents the least square analytical coefficient of the chlorophyll a concentration of the second phylum, phytoplankton; represents the average of all spectra in the canonical spectral library of phytoplankton from the third to the last phyla (Chlorophyta, Bacillariophyta, Dinophyta, and Xanthophyta); represents the linear regression analytical coefficient of the chlorophyll a concentration of phytoplankton of the corresponding phylum; represents the minimum iteration error.

[0008] Step 3: And so on; Until the least squares optimization of all the standard spectra in the standard spectrum library of the last category of phytoplankton is completed. Finally, the best standard spectrum combination of forward stepwise local optimization is obtained. , initial solution combination of chlorophyll a concentration of different phyla of phytoplankton , and the forward minimum fit residual .

[0009] Step 4: Analyze the unknown living three-dimensional fluorescence spectrum of phytoplankton using the non-negative least squares method. During the analysis, under the condition that the standard spectra of phytoplankton of other phyla except the penultimate phylum (Dinophyta) are positive stepwise local optimal standard spectra, the standard spectrum library of phytoplankton of the penultimate phylum is optimized again: , , in: It represents the three-dimensional fluorescence difference spectrum of living phytoplankton after deducting the living fluorescence intensity of the penultimate phylum (Xanthophyta); It means to traverse the least squares normalized spectrum in the normalized spectrum library of the second-to-last phylum (Dinophyta) again; represents the corresponding least square analytical concentration of phytoplankton; It represents the positive stepwise local optimal canonical spectrum of phytoplankton in other phyla (Cyanobacteria, Cryptophyta, Chlorophyta and Bacillariophyta) except the last two phyla; represents the minimum iteration error.

[0010] Step 5: Subtract the fluorescence intensity of the penultimate phylum (Dinophyta) to obtain the difference spectrum to be analyzed Under the condition that the standard spectra of the penultimate and second categories of phytoplankton are the reverse local optimal standard spectra, and the standard spectra of the third to last categories of phytoplankton are the positive local optimal standard spectra, the difference spectrum Perform the least squares analysis again: , , in: It means to traverse the least squares normalized spectrum in the normalized spectrum library of the third last phylum (Biatom) again; represents the corresponding least square analytical concentration of phytoplankton; represents the forward stepwise local optimal canonical spectrum of phytoplankton of Cyanobacteria, Cryptophyta, and Chlorophyta; represents the minimum iteration error.

[0011] Step 6: And so on; Until the least squares optimization of all the standard spectra in the standard spectrum library of the last category of phytoplankton is completed. Finally, the best standard spectrum combination of reverse step-by-step local optimization is obtained. , initial solution combination of chlorophyll a concentration of different phyla of phytoplankton , and the forward minimum fit residual .

[0012] Step 7: Compare the results of the two analyses. If there are termination conditions for any type of phytoplankton: , If established, then output If the termination condition is not met, the phytoplankton sorting method is exchanged and steps one to six are repeated using this method until the termination condition is met and the parsing result is output.

[0013] Beneficial effects: The present invention proposes an aliasing spectrum analysis algorithm based on bidirectional step-by-step local optimization. During analysis, the present invention gradually performs local optimization on the unknown aliased three-dimensional fluorescence spectrum to be analyzed in the forward analysis order, and obtains an initial solution combination; then, based on the result of the forward analysis, the reverse analysis order is gradually performed to perform local re-optimization, and a reverse re-analysis solution combination is obtained; the two groups of forward and reverse local optimal solutions are compared to determine whether the termination condition is met; if not, the algae analysis order is changed and re-analyzed; if yes, the cycle is terminated and the optimal analysis result is directly output. The bidirectional step-by-step optimization analysis algorithm described in the present invention can greatly save the amount of calculation and significantly improve the efficiency of spectral analysis on the basis of ensuring the reliability of the measurement results. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a logical schematic diagram of the present invention; Figure 2 This is a schematic diagram of the initial analysis of the forward sequence; Figure 3 It is a schematic diagram of reverse feedback re-analysis when the local optimal solution is gradually obtained in the forward direction; Figure 4Schematic representation of the mean relative error measured for the test set pure algae samples. DETAILED DESCRIPTION

[0015] The specific implementation mode of the present invention is described in detail below. This embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation mode and a specific operation process are given, but the protection scope of the present invention is not limited to the following embodiments.

[0016] Example 1 1. Establishment of interpolation standard spectral library of different types of phytoplankton based on non-uniform interpolation method: Step 1: Obtain the fluorescence spectrum of living phytoplankton, with an excitation wavelength range of 370 nm to 650 nm and an emission wavelength range of 600 nm to 720 nm. Measure the standard chlorophyll a concentration of the living phytoplankton; Step 2: Dividing the living phytoplankton fluorescence spectrum obtained in step 1 by the standard chlorophyll a concentration of the living phytoplankton to obtain a concentration normalized spectrum of the phytoplankton, i.e., a standard spectrum; Step 3: Calculate the mean fluorescence spectrum M and the corresponding relative standard deviation S of the phytoplankton according to the standard spectrum in step 2; Step 4: Calculate M±2 S, to obtain the upper and lower limits of the interpolation range of the phytoplankton standard spectrum library; Step 5: Perform uniform interpolation within the interpolation range of the phytoplankton standard spectrum library described in step 5, and the number of interpolations is 30; Step 6: Repeat steps 1 to 5 to obtain the standardized spectral library of different types of phytoplankton; 2. Obtain the living three-dimensional fluorescence spectrum of the unknown phytoplankton sample, and the spectral range is the same as step 1 in 1.

[0017] 3. The present invention analyzes the spectrum based on the least squares multivariate linear regression analysis concept.

[0018] The idea of ​​multivariate linear regression analysis is briefly described as follows: For a certain type of algae, its three-dimensional fluorescence spectrum measurement value is a fluorescence intensity data matrix, theoretically: , Among them, Mi is the in vivo three-dimensional fluorescence measurement spectrum of a certain type of algae sample, ai represents the concentration of the algae sample of this type, and fi represents the standard fluorescence spectrum of the algae sample of this type.

[0019] When there are multiple components in the sample, within the range of Lambert-Beer's law, the algae fluorescence spectrum satisfies the linear additivity, that is: , , Where: M represents the aliased in vivo three-dimensional fluorescence measurement spectrum of unknown algae samples; fij is the jth standard spectrum in the standard spectrum library of phytoplankton of the i-th phylum; ai is the chlorophyll concentration of the i-th phylum, χ is the error; λmn is the fluorescence intensity of the i-th phylum algae at the excitation wavelength n and the emission wavelength m.

[0020] In order to minimize the error, the non-negative least squares method is used to analyze the spectrum to be tested. At the same time, since the concentration of algae in the mixed algae sample to be tested will not be negative, it is necessary to add non-negative restrictions in the iterative operation and output the least squares result.

[0021] , However, when the capacity of the iterative normed spectral library of different classes of phytoplankton is large, if a global traversal method is used, it is easy to cause an explosion in the amount of computation, and the spectral analysis algorithm is also difficult to quickly converge to the true value of the mixed sample.

[0022] like Figure 1 As shown in the figure, based on the above problem, a feedback analysis algorithm for aliased fluorescence spectra of phytoplankton based on bidirectional local optimization is needed. During analysis, the unknown aliased three-dimensional fluorescence spectrum to be analyzed is gradually optimized locally in the forward analysis order, and the initial solution combination is obtained; then, based on the result of the forward analysis, the reverse analysis order is gradually optimized locally, and the reverse re-analysis solution combination is obtained; the two groups of local optimal solutions of the forward and reverse are compared to determine whether the termination condition is met; if not, the algae analysis order is changed and re-analyzed; if yes, the cycle is terminated and the optimal analysis result is directly output.

[0023] 1. Forward order initial analysis, such as Figure 2 As shown: Step 1: Analyze the three-dimensional fluorescence spectrum using the non-negative least squares method. When analyzing, under the condition that the standard spectrum of phytoplankton from the second to the last phylum of phytoplankton is the average value of all spectra in the standard spectrum library, the standard spectrum library of the first phylum of phytoplankton is optimized in the entire library: , in: Represents the three-dimensional fluorescence spectrum of unknown phytoplankton to be analyzed; represents the average of all spectra in the canonical spectral library of phytoplankton from the second to the last phylum of phytoplankton (Cryptophyta, Chlorophyta, Bacillariophyta, Dinophyta, and Xanthophyta); The linear regression analytical coefficients representing the chlorophyll a concentration of the second to last phylum of phytoplankton; It means to traverse the minimum residual standard spectrum in the standard spectrum library of phytoplankton of the first phylum (Cyanobacteria); represents the least square analytical coefficient of the chlorophyll a concentration of the first phylum, phytoplankton; represents the minimum residual.

[0024] Step 2: Analyze the three-dimensional fluorescence spectrum of the living body Deduct the fluorescence intensity of the first phylum (Cyanobacteria) from the equation to obtain the difference spectrum to be analyzed. ; Under the condition that the standard spectrum of the first phylum (cyanobacteria) of phytoplankton is the minimum residual standard spectrum, and the standard spectrum of the third to the last phylum of phytoplankton is the average value of all spectra in the standard spectrum library, the difference spectrum to be analyzed is analyzed by the non-negative least squares method: , , in: It represents the minimum residual normalized spectrum in the normalized spectrum library of phytoplankton of the second phylum (Cryptophyta); represents the least square analytical coefficient of the chlorophyll a concentration of the second phylum, phytoplankton; represents the average of all spectra in the canonical spectral library of phytoplankton from the third to the last phyla (Chlorophyta, Bacillariophyta, Dinophyta, and Xanthophyta); represents the linear regression analytical coefficient of the chlorophyll a concentration of phytoplankton of the corresponding phylum; represents the minimum iteration error.

[0025] Step 3: And so on; Until the least squares optimization of all the standard spectra in the standard spectrum library of the last category of phytoplankton is completed. Finally, the best standard spectrum combination of forward stepwise local optimization is obtained. , initial solution combination of chlorophyll a concentration of different phyla of phytoplankton , and the forward minimum fit residual .

[0026] 2. When the local optimal solution is gradually obtained in the forward direction, reverse feedback is used for further analysis, such as Figure 3 As shown: including steps 4 to 7: Step 4: Analyze the unknown phytoplankton living three-dimensional fluorescence spectrum to be analyzed by non-negative least squares method. During the analysis, under the condition that the standard spectra of phytoplankton of other phyla except the penultimate phylum (Dinophyta) are positive stepwise local optimal standard spectra, the standard spectrum library of phytoplankton of the penultimate phylum is optimized again: , , in: It represents the three-dimensional fluorescence difference spectrum of living phytoplankton after deducting the living fluorescence intensity of the penultimate phylum (Xanthophyta); It means to traverse the least squares normalized spectrum in the normalized spectrum library of the second-to-last phylum (Dinophyta) again; represents the corresponding least square analytical concentration of phytoplankton; It represents the positive stepwise local optimal canonical spectrum of phytoplankton in other phyla (Cyanobacteria, Cryptophyta, Chlorophyta and Bacillariophyta) except the last two phyla; represents the minimum iteration error.

[0027] Step 5: Subtract the fluorescence intensity of the penultimate phylum (Dinophyta) to obtain the difference spectrum to be analyzed Under the condition that the standard spectra of the penultimate and second categories of phytoplankton are the reverse local optimal standard spectra, and the standard spectra of the third to last categories of phytoplankton are the positive local optimal standard spectra, the difference spectrum Perform the least squares analysis again: , , in: It means to traverse the least squares normalized spectrum in the normalized spectrum library of the third last phylum (Biatom) again; represents the corresponding least square analytical concentration of phytoplankton; represents the forward stepwise local optimal canonical spectrum of phytoplankton of Cyanobacteria, Cryptophyta, and Chlorophyta; represents the minimum iteration error.

[0028] Step 6: And so on; Until the least squares optimization of all the standard spectra in the standard spectrum library of the last category of phytoplankton is completed. Finally, the best standard spectrum combination of reverse step-by-step local optimization is obtained. , initial solution combination of chlorophyll a concentration of different phyla of phytoplankton , and the forward minimum fit residual .

[0029] Step 7: Compare the results of the two analyses. If there are termination conditions for any type of phytoplankton: , If established, then output If the termination condition is not met, the phytoplankton sorting method is exchanged and steps one to six are repeated using this method until the termination condition is met and the parsing result is output.

[0030] Example 2 Three-dimensional fluorescence spectra of living phytoplankton under different habitat conditions Common phytoplankton species in my country's coastal and freshwater waters were selected, including Cyanobacteria, Chlorophyta, Bacillariophyta, Dinophyta, Xanthophyta and Cryptophyta, see Table 1. They were cultured according to GB / T21805-2008 "Chemicals - Algal Growth Inhibition Test".

[0031] Table 1 Experimental sample types

[0032]

[0033] Take an appropriate amount of algae sample mother solution at different growth stages, dilute it according to a certain volume ratio, and measure the three-dimensional fluorescence spectrum of the sample in vivo with a fluorescence spectrophotometer. The excitation wavelength range is 370 nm to 650 nm, and the emission wavelength range is 600 nm to 720 nm. At the same time, refer to the national standard method to measure the standard concentration of chlorophyll a in the algae sample.

[0034] Example 3 Obtaining interpolation standard spectral libraries of different phytoplankton The algae samples with known chlorophyll a standard concentration obtained in Example 2 were divided into an experimental set and a test set in a ratio of 7:3.

[0035] The three-dimensional fluorescence spectrum of the living algae in the experimental set was divided by the standard concentration of chlorophyll a, thereby obtaining a series of three-dimensional fluorescence standard spectra of living algae. According to the standard spectra of different categories, the mean fluorescence spectrum M and the corresponding relative standard deviation S of the phytoplankton were calculated; M ± 2 S, obtaining the upper and lower limits of the interpolation range of the standard spectrum library of phytoplankton; performing uniform interpolation within the interpolation range of the standard spectrum library of phytoplankton described in step 4, the number of interpolations is 10, thereby establishing the standard spectrum library of phytoplankton.

[0036] The above method obtains the standardized spectral library of phytoplankton of other specific phyla including Cyanobacteria, Chlorophyta, Bacillariophyta, Dinophyta, Xanthophyta and Cryptophyta.

[0037] Example 4 Comparison of hierarchical and partitioned analysis effects The analytical method of the present invention was used to analyze the living fluorescence spectra of the phytoplankton in the test set obtained in Example 2, and the analytical results were compared with the standard chlorophyll a concentration of the sample and the direct analytical results of the least squares multivariate linear regression algorithm (i.e., the common analytical algorithm) of the non-stratified partition. The test set samples included 309 pure algae samples and 28 mixed samples. and relative measurement error Indicates the parsing accuracy of the algorithm.

[0038] , , in: and They represent the total number of samples of a certain phylum of phytoplankton and the number of samples misidentified by analysis. Misidentification by analysis means that the dominant population is analyzed incorrectly, such as pure diatom samples being misidentified as other pure non-diatom samples or mixed algae samples with non-diatoms as the dominant population; and They respectively represent the standard concentration and analytical concentration of chlorophyll a in a certain phylum of phytoplankton samples.

[0039] Table 2 The number and rate of misidentification of the test set samples by the bidirectional stepwise optimization analysis algorithm

[0040]

[0041] Table 2 shows that the bidirectional stepwise optimization analysis algorithm described in the present invention has an accuracy rate of 100% for pure samples of Cyanobacteria, Chlorophyta, Bacillariophyta and Cryptophyta, and the number of misidentifications for pure samples of Dinophyta and Xanthophyta are 1 and 5 respectively, and the misidentification rates are 1.7% and 8.7% respectively. There is one misidentification of mixed samples of Bacillariophyta and Dinophyta, and the misidentification rates are 5.9% and 11.1% respectively. The overall misidentification rate of pure samples is 1.9% (6 / 309), and the overall misidentification rate of mixed samples is 7.1% (2 / 24), indicating that the bidirectional stepwise optimization analysis algorithm can accurately identify most algae samples, and only a small number of misidentifications are found for Bacillariophyta, Dinophyta and Xanthophyta with high similarity fluorescence spectra.

[0042] Figure 4 The results show that the average measurement errors of the bidirectional stepwise optimization analysis algorithm for pure Cyanobacteria, Chlorophyta, Bacillariophyta, Dinophyta and Cryptophyta algae samples are 7.6%, 16.3%, 15.9%, 9.6% and 5.2%, respectively, all below 20%; the average measurement error for pure Xanthophyta samples is 27.9%. The average measurement error for mixed samples of algae samples of different phyla is slightly higher overall, but still basically below 30%.

[0043] Table 3 Average sample processing data volume and analysis time of the test set

[0044]

[0045] Table 3 shows that in this case, when the global traversal method of all algae standard spectral libraries is adopted, the average amount of data that needs to be processed when the least squares analysis algorithm analyzes the spectrum is 7.25. 10 8, the average spectrum analysis time is 36619.2 s; when the analysis algorithm described in the present invention is used, the average amount of data to be processed when analyzing the spectrum is 87000, the required spectrum analysis time is 3.18 s, and the calculation speed is increased by nearly 11515 times. This shows that the bidirectional step-by-step optimization analysis algorithm described in the present invention can greatly save the amount of calculation and significantly improve the efficiency of spectrum analysis on the basis of ensuring the reliability of the measurement results.

[0046] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An aliasing spectrum analysis algorithm based on bidirectional stepwise local optimization, characterized in that: The steps include: Step 1: Analyze the three-dimensional fluorescence spectrum using the non-negative least squares method. When analyzing, under the condition that the standard spectra of the second to the last phylum of phytoplankton are the average values ​​of all spectra in the standard spectrum library, the standard spectrum library of the first phylum of phytoplankton is optimized in the entire library: ， in: Represents the three-dimensional fluorescence spectrum of unknown phytoplankton to be analyzed; represents the average of all spectra in the standardized spectral library of phytoplankton from the second to the last phylum of phytoplankton; The linear regression analytical coefficients representing the chlorophyll a concentration of the second to last phylum of phytoplankton; It means to traverse the minimum residual standard spectrum in the standard spectrum library of the first phylum phytoplankton; represents the least square analytical coefficient of the chlorophyll a concentration of the first phylum, phytoplankton; represents the minimum residual; Step 2: Analyze the three-dimensional fluorescence spectrum of the living body Deduct the fluorescence intensity of the first category from the equation to obtain the difference spectrum to be analyzed ; Under the condition that the standard spectrum of the first category of phytoplankton is the minimum residual standard spectrum, and the standard spectrum of the third to the last category of phytoplankton is the average value of all spectra in the standard spectrum library, the difference spectrum to be analyzed is analyzed by the non-negative least squares method: ， ， in: It represents the minimum residual standard spectrum in the standard spectrum library of phytoplankton of the second phylum; represents the least square analytical coefficient of the chlorophyll a concentration of the second phylum, phytoplankton; represents the average of all spectra in the standard spectral library of phytoplankton from the third to the last phylum; represents the linear regression analytical coefficient of the chlorophyll a concentration of phytoplankton of the corresponding phylum; represents the minimum iteration error; Step 3: Repeat this process until the least squares optimization of all the standard spectra in the standard spectrum library of the last phylum of phytoplankton is completed, and finally the best standard spectrum combination of forward stepwise local optimization is obtained. , initial solution combination of chlorophyll a concentration of different phyla of phytoplankton , and the forward minimum fit residual ; Step 4: Analyze the unknown living three-dimensional fluorescence spectrum of phytoplankton by non-negative least squares method. When analyzing, under the condition that the standard spectra of phytoplankton of other categories except the penultimate category are positive stepwise local optimal standard spectra, the standard spectrum library of phytoplankton of the penultimate category is optimized again: ， ， in: It represents the three-dimensional fluorescence difference spectrum of living phytoplankton after deducting the fluorescence intensity of the penultimate category; It means to traverse the least squares normative spectrum in the normative spectrum library of the second to last category again; represents the corresponding least square analytical concentration of phytoplankton; It represents the positive stepwise local optimal normative spectrum of phytoplankton except the last two classes; represents the minimum iteration error; Step 5: Subtract the fluorescence intensity of the second-to-last category to obtain the difference spectrum to be analyzed Under the condition that the standard spectra of the penultimate and second categories of phytoplankton are the reverse local optimal standard spectra, and the standard spectra of the third to last categories of phytoplankton are the positive local optimal standard spectra, the difference spectrum Perform the least squares analysis again: ， ， in: It means to traverse the least squares normative spectrum in the third last category normative spectrum library again; represents the corresponding least square analytical concentration of phytoplankton; represents the forward stepwise local optimal canonical spectrum of phytoplankton of Cyanobacteria, Cryptophyta, and Chlorophyta; represents the minimum iteration error; Step 6: Repeat this process until the least squares optimization of all the standard spectra in the standard spectrum library of the last category of phytoplankton is completed, and finally the best standard spectrum combination of reverse stepwise local optimization is obtained. , initial solution combination of chlorophyll a concentration of different phyla of phytoplankton , and the forward minimum fit residual ; Step 7: Compare the results of the two analyses. If there are termination conditions for any type of phytoplankton: ， If the termination condition is not met, then the phytoplankton sorting method is exchanged, and steps one to six are repeated using this method until the termination condition is met and the analysis result is output.

Images