Design And Implementation Of Cloud Platform For Water Quality Detection Based On UV-Vis Spectroscopy

With the steady progress of my country's social modernization process,the protection of the natural environment has become an important topic,and the state's efforts to protect the natural ecology have increased year by year.The"14th Five-Year Plan"also puts forward higher requirements for the protection of the water environment.For"green"development,it is necessary to strengthen ecological restoration on the basis of ecological protection.As a branch of environmental protection,water quality monitoring is particularly important.Using the water quality monitoring cloud platform to assist water quality monitoring is the only way to improve monitoring efficiency.Although the traditional chemical water quality detection method has the advantage of high detection accuracy,its detection needs to be carried out in a laboratory environment,and the detection process is cumbersome,the detection time is long,and there are defects such as secondary pollution.Nowadays,UV-Vis spectroscopy is widely used in the field of water quality detection.It has the advantages of fast detection speed,real-time in-situ detection,simple operation,and no pollution in the whole detection process,which better makes up for the shortcomings of traditional methods.To realize the control and monitoring of water quality,it is not only necessary to configure a more accurate algorithm model,but also to be supported by a considerable amount of data,so that the monitoring model can be more accurate and more credible.Therefore,this paper proposes the design and implementation of a cloud platform for water quality monitoring based on UV-Vis spectroscopy.Based on this,this paper is jointly funded by the National Natural Science Foundation of China Youth Fund Project(61805029)and the Chongqing Municipal Education Commission Science and Technology Research Project(KJ1709201)to achieve efficient,real-time and in-situ monitoring of the water environment.The design and implementation of the water pollution cloud platform based on UV-Vis spectroscopy.The main research work is as follows:1)Construction of a scientific research-grade water quality spectral data acquisition system based on UV-Vis spectroscopy.Using Ocean Optics'DH2000deuterium halogen tungsten HAMAMATSU company's C10082CA spectrometer to build a water quality data acquisition system,through the actual water samples,high-purity water and according to"Water Quality-Determination of Permanganate"(GB11892-89)and"Water Quality-Determination of Permanganate"The standard solution of sodium Na₂C₂O₄ prepared in accordance with the Determination of Salts"(ISO 8467-1986),to obtain UV-Vis spectral data of water samples.2)Construction of a cloud platform for water quality monitoring based on UV-Vis spectroscopy.By selecting the ATHB300 UV-Vis spectral data sensing probe assembled by OPTOSKY's ATG1030 pulsed xenon lamp module and ATP2000 spectrometer,the real-time monitoring and data collection of water quality data are performed,and the ALIENTEK company's ATK-M751 4G-DTU,to transmit water quality data.Alibaba Cloud servers are used to build a cloud platform to analyze the spectral data of water quality,and to monitor the quality of the water environment in the covered area in real time,providing effective basis and data support for judging the discharge of industrial wastewater in key river basins and the supervision and protection of the ecological water environment.3)Research on COD classification of industrial wastewater based on neural network algorithm.Aiming at the shortcomings of using convolutional neural network(CNN)or long short-term memory network(LSTM)independently to establish a COD classification model for industrial wastewater,a hybrid model based on CNN and LSTM was constructed.Wastewater COD data,and then use CNN to extract abstract features from the data,and finally input them into LSTM to obtain the result of classifying industrial wastewater according to COD concentration.The classification accuracy of the model reaches 98.66%,which is higher than using CNN and LSTM alone.3.36%and 4.7%.