Design Of COD Detecting Device Based On UV-Near Infrared Spectroscopy

Traditional method of measuring Chemical Oxygen Demand(COD)has various disadvantages such as high cost of chemicals,low efficiency of detection,difficulty on continuous inspection,second pollution of the waste.Considering this problem,based on the existing ultraviolet and near-infrared spectrum COD measuring technique,investigated on the feasibility of predicting COD in waste according to the information contained in ultraviolet and near-infrared spectrum.This essay demonstrates a design of a device that automatically samples the water and calculates the composition of COD based on the spectrum absorbed by water.This device is highly efficient and fairly accurate.It is also able to exchange messaged conveniently with users via graphical interfaces.Moreover,it is possible to widely apply this device in industrial process.In order to accomplish the investigation goal,this essay is mainly working on:1.This essay firstly introduces and analyzes domestic and overseas research results of water quality supervision,then combines them with traditional COD measuring method and existing spectrum measuring technique.The feasibility and practical values of using spectrum water sample to directly calculate COD composition is also discussed.This brings research and design of automatic water sample testing device on track.2.The experiment collected water samples from 100 factories.Each sample is divided into two sets: one set is tested on spectrum of wavelength in 190-2200 nm,this is the experiment group;the other set is tested using traditional method,this is the control group.Then Partial least squares(PLS)is used on the absorbed spectrum to build the COD predicting model.As a result,the relativity of predicted value and actual value is 0.995,standard error is9.54mg/L,and it is a trustworthy prediction.Similarly,using the prediction model to test on different types of wastes also obtains satisfying results.3.An optic COD measuring device is designed,which contains sampling system,optic system and embedded system.In accordance to the process of manual measuring and modularize its function,a continuous automatic running sampling system is designed;As required by the prediction model,equipment of optic system such as light source and sensors are selected;Based on controlling performance and design costs,an embedded system with an ARM chip in core is designed.Under control of embedded system,sampling system and optic system can run automatically in procedure planned.4.An Android application based on prediction algorithm and hardware working procedure is compiled.The main function of the application is to process system data and exchange information with users via graphic interface.This application uses I/O gates of embedded system to read spectrum data and device status,after calculation and transformation,it saves the data into storage elements.The application also provides touch screen interface,on which the users can check the absorbed spectrum,COD composition and working status of sampling devices.Besides,users can switch to manual mode via menu options.