Study On Theory And Application Of Algae Recognition Based On Fluorescence Method

Phytoplankton in the sea is the main primary producer and energytransducer in ocean ecological system. By the recognition of algae that is themain ocean phytoplankton, we can estimate community structure and energydistribution in the sea area and realize monitoring and governing oceanpollution. Based on the analysis and synthesis of principle of fluorescencemeasurement and theory of artificial neural network, this paper proposes amethod that the single algae could be recognized based on fluorescencemethod and BP network, applies it to recognition of three kinds of single algae,and designs systematic scheme which is so called optical fiber andfluorescence system for automatic algae recognition. The main researchcontents are as followings: (1) Based on the fundamental theory of fluorescence measurement, theprinciple of the algae fluorescence detection is studied and the feasibility offluorescence method is demonstrated in order to provide foundation for thereasonable online recognition of different kind of algae. (2) Not only are the absorption spectra, excitation spectra and fluorescenceemission spectra properties are analyzed by the experiments, but also thecharacteristic parameters are picked-up for algae recognition and bestexcitation wavelength and fluorescence detection wavelength are selected. (3) The total scheme of optical fiber and fluorescence system for automaticalgae recognition are based on PC. This system includes the double-channelsmeasurement technology for weakening the influence of instabilitylamp-house, the probe for effective collecting fluorescence without scatteredlight, low noise preamplifier average detection technique in time field to pickup the weak fluorescence from the measured signals with noise. (4) Based on fluorescence spectra of algae, BP network model forrecognition of three kinds of algaes is designed, recognition experiment ofthree kinds of single algae has been carried with BP network, whichdemonstrates the feasibility fluorescence method and artificial neural network IIpattern recognition theory algae recognition. This research applies fluorescence method and artificial neural networkpattern recognition theory to optical fiber algae monitor system that canrealize recognition of three kinds of algaes. This system has real time datacollection, data processing and results displaying functions. A novel technicalmeans for algae state monitoring is provides by the paper.