| Total phosphorus and total nitrogen are distributed in various water environments.Too much phosphorus and nitrogen will lead to overnutrition and overgrowth of plankton,which will make the water environment stink,kill fish,and ultimately harm human beings.Among the total phosphorus and total nitrogen detection methods,Surface Enhanced Raman Spectroscopy,as a modern advanced "fingerprint" spectrum detection technology,has a fast detection speed and high sensitivity,and has a broad prospect in environmental monitoring.Based on this,the Surface Enhanced Raman Spectroscopy technique is used for the first time in the spectral detection and concentration prediction of total phosphorus and total nitrogen.First of all,the research background and significance of this topic are introduced,and the commonly used detection methods of total phosphorus and total nitrogen are investigated and their advantages and disadvantages are summarized,and the research direction and objectives are established.In addition,the enhancement mechanism of Surface Enhanced Raman Spectroscopy and the pretreatment methods of spectral analysis are described.Secondly,three kinds of reinforced substrates,namely nano silver sol,nano gold sol and silver shell gold core particles,were prepared by oxidation-reduction method.By controlling the volume of sodium citrate,sol particles with different particle sizes were prepared,and the enhancement effect on total phosphorus and total nitrogen was observed,Thirdly,the total phosphorus and total nitrogen of a series of concentration gradients were measured,the vibration modes of the characteristic peaks of the surface enhanced Raman spectrum are analyzed,one of the characteristic peaks is selected as a resear ch object,the peak intensity of different concentrations is taken as an input quantity,46 samples and 23 samples are combined with particle swarm optimization(PSO)to optimize the BP neural network algorithm to respectively establish concentration prediction models of total phosphorus and total nitrogen,8 samples are taken as test sets to detect the models,and model evaluation is carried out according to correlation coefficient,root mean square error,relative deviation and recovery rate.Finally,the real seawater environment was simulated,combined with Raman spectroscopy and Particle Swarm Optimization neural network prediction model to detect the concentration of total phosphorus and total nitrogen.On the basis of fast prediction of concentration,through comparing the changes of total phosphorus and total nitrogen concentration after 48 hours,the removal rate research experiment of total phosphorus and total nitrogen was designed. |
