Research On Measurement Method Of Planktonic Algae Based On Three-wavelength Fluorescence Spectrometry

The phenomenon of water bloom in rivers and lakes in China is becoming increasingly prominent,which poses a threat to human production and life.Therefore,it is urgent to establish an online method for meas:uring phytoplankton and its concentration in order to realize the daily monitoring and early warning function of phytoplanktorn.Fluorescence measurement is one of the hotspots in modern measurement.It can measure the species and concentration of substances by measuring the fluorescence spectrum of substances.It has broad application prospects.On the basis of fluorescence measurement of phytoplankton,this paper proposes a method of phytoplankton identification and concentration measurement based on three wavelengths by analyzing the characteristics of photosynthetic pigments of phytoplankton and experimental research.At the same time,an implementation scheme of phytoplankton identification and measurement device is designed.Based on the analysis of common algae in freshwater waters of China,this paper selects three species of algae:Chlorophyta,Cyanophyta and Diatom,as the main research objects,and uses the fluorescence spectrum data of three pure algae as training sets.It is used to establish the algae measurement model;the fluorescence spectrum data of the mixed samples of pure algae and nursing school river water,West Lake water,and community sewage water are used as test sets to verify the accuracy of the model and achieve the level of the class.The measurement of phytoplankton is achieved.The specific research is as follows:(1)Through the analysis of the characteristics of photosynthesis pigments and absorption spectra of planktonic algae,it is found that the photosynthetic pigments in different species of algae have certain differences,and the species of photosynthetic pigments in the same species of algae are the same,resulting in the same species of algae.The fluorescence excitation spectra are very similar,and the normalized fluorescence excitation spectra of different species are significantly different,which constitutes the main basis for classification of planktonic algae.(2)According to the absorption spectrum characteristics of the selected photosynthetic pigments of three species of algae,the four excitation wavelengths of 440nm,470nm,530nm and 630nm were selected for experimental testing,and the gradient boost decision tree(GBDT)was used for wavelength screening.The root node of each tree in GBDT represents a feature.For the algae fluorescence data of this paper,each excitation wavelength represents a characteristic,and the fluorescence value corresponding to each excitation wavelength represents a characteristic value,and the characteristic value is tested according to the test result.Classify the samples.In this way,the feature with high contribution to the classification result can be judged from the classification result,thereby achieving the purpose of feature screening.In this paper,the GBDT algorithm is used to screen the selected four excitation wavelengths.After removing the excitation wavelength of 440nm,a three-wavelength-based method for identification and co,ncentration measurement of phytoplankton is proposed.(3)Established a model of phytoplankton identification.Using the spectral data of pure algae,the identification model of three species of cyanobacteria,cyanobacteria and diatoms was established by SVM algorithm,and the mixed samples of river water,West Lake water,sewage and pure algae were used to verify the identification.The accuracy of the model,the average accuracy of the cyanobacteria,green algae and diatom recognition models are 92.1%,81.0%and 85.2%,respectively.(4)Established a model for measuring the concentration of phytoplankton.Based on the spectral data of pure algae,the regression model between the fluorescence intensity and algae concentration of Chlorophyta and cyanophyta was established,and the accuracy of the regression model was evaluated by regression evaluation index.The green algae regression model was 0.8821.The diatom regression model is 0.9590.(5)A scheme based on a three-wavelength algae fluorescence measuring device was designed.The device is based on three characteristic wavelength highlighting LEDs and fluorescent detection probe modules,and the outer casing of the optical path structure is designed.The printing material of the PLA material ensures the strength of the whole structure,and can be placed in the mobile water equipment to realize the Dynamic measurement of the specific water area.