Water Quality COD Detection Method And System Development Based On Fluorescence Spectroscopy

Water is an essential resource for human survival.Since the beginning of the 21 st century,the problem of water pollution in China has become increasingly serious,with frequent incidents of water pollution that have seriously endangered the health of the population.Therefore,effective monitoring of the national water environment has become important.Chemical Oxygen Demand(COD)is an indicator that characterizes the content of organic matter in water and is one of the key pollution parameters monitored by the state.In this paper,based on fluorescence spectroscopy,a measurement model between fluorescence spectroscopy and water quality COD is established,and a water quality COD detection system based on fluorescence spectroscopy is designed and implemented.The main contents of this thesis are:1.The design of a water quality COD detection system was completed.Based on the application scenario of the system,the requirements of the components in the system were determined,and a hardware design scheme with a CCD portable spectrometer as the system detector and a semiconductor laser with a wavelength of 405 nm as the light source was proposed.The structure of the optical path commonly used in fluorescence systems was optimised to simplify the optical path and improve the immunity of the system optical path.2.A water quality COD prediction model based on fluorescence spectroscopy is developed.Based on the proposed hardware design scheme,a temporary experimental platform was built and the fluorescence spectra of the actual water samples were collected based on the experimental platform.Using the collected fluorescence spectra,a Partial Least Squares(PLS)-based water quality COD prediction model was established,and the Principal Component Analysis(PCA)and the Successive Projections Algorithm(SPA)were used.The final results show that the modelling effect of PCA is better than that of SPA-based feature extraction.Moreover,when the number of PCA principal components is 15,the PLS model is the best;at this time,the coefficient of determination of the test set is 0.9581 and the root mean square error is0.1685 mg/L,which indicates that the model had a high prediction accuracy and proved the feasibility of the design solution.3.A fluorescence spectroscopy-based water quality COD detection system was implemented,and the system was tested.The control system of the water quality detection system was designed and developed for completing the control of the system components.218 water samples were collected to test the performance of the system,of which 138 water samples were used for the training of the model and 80 water samples were used to test the prediction performance of the model.The test results show that the average error of the system is 9.84% for the 80 test water samples,which meets the requirements in China’s "Automatic Surface Water Monitoring Specification".