【作者】 史慧超；

【导师】 王玉田；

【作者基本信息】 燕山大学 ， 测试计量技术及仪器， 2010， 硕士

【摘要】 海洋中的浮游植物是海洋生态系统中能量的主要转换者和最主要的生产者,海藻则是海洋中主要的浮游植物,检测海洋中海藻的种类和浓度,可以估测海域生态系统的群落结构和能量分布状态,实现对海洋污染的监测和治理。本文根据荧光测量机理和人工神经网络理论,提出了将荧光法与BP神经网络相结合的单一海藻种类识别和浓度测量方法。通过分析荧光测量的基本原理、特点、海藻荧光检测机理,研究不同种类、不同浓度海藻溶液及其所含特征色素的吸收光谱、激发光谱、荧光光谱特性,提取海藻种类识别和浓度测量的特征参量,为设计光纤荧光测量系统提供了理论和实验依据。提出光纤荧光测量系统的总体方案,设计测量系统的光路部分,其中包括:光源的选择及驱动电路的设计,滤光及光的传输的实现,光耦合方法及结构设计等;设计微弱荧光信号检测方案,其中包括:光电二极管的选择及电路设计,光电倍增管的选择及电路设计,微弱信号检测技方法的实现。在对荧光法检测海藻和人工神经网络理论进行分析的基础上,应用BP网络进行三种不同门类海藻的识别,并对相应种类海藻进行浓度预测。设计海藻种类识别BP网络及海藻浓度预测BP网络。对三种不同门类单一海藻的种类识别和浓度预测实验结果进行分析,验证神经网络应用于光纤荧光海藻检测的方法的可行性。本研究将荧光法和人工神经网络模式识别和参数预测理论相结合应用于光纤海藻测量系统,能对三种不同门类的海藻进行种类识别和浓度测量,具有数据实时采集、处理及结果显示等功能,为海藻状态监测提供了一种新手段。更多还原

【Abstract】 Phytoplankton in the sea is the main energy transducer and producer in ocean ecological system, and algae are the major marine phytoplankton. By recogning algae and measuring its concentration, we can estimate community structure and energy distribution in the sea area and realize monitoring and governing ocean pollution.This paper proposes a combination of fluorescence method and BP neural networks to realize single algal species identification and concentration measurement. By analyzing the basic principles and characteristics of fluorescence measurement and algae fluorescence detection mechanism, studying the different types and different concentrations solution of seaweed and characteristics of pigments contained in the absorption spectra, excitation spectra, fluorescence spectral characteristics,extract characteristic parameter for algae species identification and concentration measurement ,provide theoretical and experimental basis for the design of optical fiber fluorescence measurement system.Propose total designation scheme of the optical fiber fluorescence measurement system,design optical part of the measurement system, including: the choice of light source and drive circuit design, the realization of filter and optical transmission, optical coupling method and structural design, etc.;design weak fluorescent signal detection method, including: the choice of photodiode and circuit design, the choice of photomultiplier tubes and circuit design, the realization of weak signal detection.Based on analysing the theory of fluorescence detection and artificial neural network, apply BP network to recognize three kinds of algae, and then measure its concentration,design algal species identification BP network and algae concentration predicted BP network Analysis the categories identification and concentration estimation results of three different kinds of algae, verify the feasibility of this method.This research applies fluorescence method and artificial neural network pattern recognition and Parameter estimation theory to optical fiber algae monitor system that can realize recognition and concentration estimation of three kind of algae. This system has real time data collection, data processing and results displaying functions. A novel technical means for algae state monitoring is provides by the paper.更多还原