Based On The Pls Water Quality Study Of The Multi-parameter Monitor

At present, automatic on-line water quality monitor most of the design principle is based on national standard method, is about the traditional chemical method of analysis of the process online. Analysis of NO3-, COD, and TOC content is usually by adding chemical reagents and then measuring their response to the indirect calculation of the concentration of product, or heating, burning and other methods to measure the content of CO₂ generated by indirectly measuring the concentration of COD or TOC. These methods are not only a waste of chemical reagents, waste energy, time-consuming, but also to bring about secondary pollution, such as water quality. Measurement parameters and a single type of comparison, if a number of parameters to be measured, it must purchase more than one monitor to meet the requirements. This is not only the purchase costs, maintenance costs, but also measurement procedures, and repetitive operation inconvenience. Therefore, the development of simple, fast, reliable, no secondary pollution, and measurement of multiple parameters at the same time automatic online water quality monitoring is very important. Existing research shows that the decision to waste water NO₃-, COD, TOC, ABS254 value of the size of the pollutants in areas with strong UV absorption. Under certain conditions, these pollutants and the absorbance of NO3-, COD, TOC, ABS254 have a very good correlation between the use of such a correlation can be used to direct determination of UV absorbance of NO₃-, COD, TOC, ABS254. No need for chemical reagents to determine the fast, easy to use. In this paper, the UV absorption spectra of organic matter to start, choose a certain representation of some of the experimental chemical reagents to explore the UV absorbance and concentration of these reagents, the use of partial least squares regression model, the establishment of the direct determination of UV absorption spectrum of water NO₃-, COD, TOC methods and the use of this model simulation to predict waste water samples. In this paper, the use of SIMCA-P (11.5 Demo) PLS universal platform for the completion of modeling and forecasting. The establishment of 30 samples using partial least-squares regression model to predict the effect of self-simulation of waste water, when the number of principal component extraction in more than 10, all parameters are the relative error <2.1000%; when the number of principal components extracted for the 21 o'clock five parameters are the average relative error of <0.0050%.This article is designed to develop multi-parameter water quality monitoring device is required for hardware, given the entire hardware platform, hardware schematics and introduced measures such as anti-jamming; software, mainly in the Matlab platform to achieve a partial least squares computation of the main features, At the same time, given the realization of PLS flow chart of the software and has done a detailed briefing.In this paper, the results of partial least-squares in-depth study of UV absorption spectrometry determination of multi-parameter provides a valuable reference, as well as to further improve and lay a foundation for this method.