Research On Intelligent Monitoring System For Ship Discharge Pollutants

It has aroused widely attention by international societies to the rising marine environment pollution due to the pollutant emissions of vessels' oily sewage and the sulfide in the exhaust gas.It plays an important role during the marine environmental protection by the effective monitoring on emission pollutants from vessels.There is non-linear error which will affect detection precision,as the traditional marine oil content meter is easily influenced by the interference factors such as bubbles during the actual detection.While currently there is not a set of completely mature system which can be accurately applied to the on-line monitoring of sulfide and other pollutants for the vessels' exhaust gas.It is completely depended on the self-consciousness of crew and shipowner to abide by the relevant convention on environmental emissions.This paper researches and designs an intelligent monitoring system on the emission pollutants from vessels faced with the current demand and new tendency of “the connected ship”development,which contains three parts: marine oil concentration detection principle prototype,on-line monitoring system on concentration of sulfur dioxide from the vessels' exhaust gas and the pc monitoring system.The paper systematically and deeply carries out the research on intelligent monitoring system of the emission pollutants from vessels with the background of combining optical detection technology and soft sensor method.Main research points:1.The development of marine oil concentration detection principle prototype.(1)The hardware design of marine oil concentration detection principle prototype.It mainly contains sensor photoelectric structure design,hardware circuit design and signal acquisition module design.(2)The research on the detection model of marine oil concentration with least squares support vector machine(LS-SVM).The article aims at the issues of traditional marine oil concentration detection based on nephelometry,the non-linear error caused by interference factor such as bubbles.Adopting LS-SVM to establish detection models of oil concentration which is best widely used in solving small sample statistics,non-linear modeling.The research results show that it can be used in the design on detection principle prototype of marine oil concentration.(3)The research on optimization of marine oil concentration detection model parameters based on LS-SVM.Aiming at the blindness of man-selected on it,and the problems of influence on model prediction accuracy.The article adopts particle swarm optimization(PSO)algorithm to optimize LS-SVM marine oil concentration detection model parameters.The RMSE and MRE are used as the evaluation standard to contrast the result of its LS-SVM and least square fit.The results show that adopt particle swarm optimization to optimize least squares support vector machine(LS-SVM)model parameters PSO-LS-SVM oil content detection model's precision more higher,and effectively avoid the decrease on model generalization performance caused by blind choosing of model parameters,which is suitable for the design of detection principle prototype on marine oil concentration.2.The paper is based on the research of de-noising algorithm on sulfur dioxide concentration detection signal by wavelet analysis,aiming at the detector signal easily being drown in the noise,sulfur dioxide's absorption to infrared light is very weak during actual measurement,even by enlarging smoothing,the signal is inevitably effected by the noise of amplifier,the external environment,radiation source.The paper did wavelet threshold de-noising to the noise signal utilizing the advantages of wavelet de-noising algorithm such as time domain and localization of frequency domain,singularity of detection signal and abrupt structure.3.The design of the host computer monitoring software of intelligent monitoring system with emission pollutants from vessels based on Lab VIEW software platform.