Design Of Water Quality Monitoring System Based On NB-IoT And Research On Data Classification

With the rapid development of the national economy,a series of environmental pollution problems will inevitably bring,among which the damage to the water environment is increasingly serious.However,both human health and social stability are closely related to the safety of the water environment.In order to do a good job in the supervision and prevention of water resources,a perfect water quality monitoring system can achieve accuracy and real-time of water quality monitoring,and provide effective technical support for solving the problem of water pollution.Therefore,the research on water quality monitoring technology is of great significance.Traditional communication technologies such as Wi-Fi,Zig Bee and Bluetooth cannot meet the requirements of remote water quality monitoring due to the short communication distance;however,communication technologies such as GPRS,4G,and 5G,due to issues such as energy consumption,signal coverage,and high traffic cost,which have stringent requirements for features such as throughput,latency,and availability,and are not suitable for application in water quality monitoring.Therefore,this paper designs a water quality monitoring system based on the Narrow-Band Internet of Things(NB-Io T),which is mainly composed of water quality data acquisition module,NB-Io T data transmission module and water quality monitoring management platform.The system can not only display water quality information to users in real time,but also classify water quality data automatically and accurately.The main work of the paper is summarized as follows:(1)Aiming at the problems of poor generality,complicated calculation and low accuracy of traditional water quality data classification methods,a classification method based on weighted Naive Bayes is proposed.This method comprehensively considers the influence of water quality attributes and their value on the classification results,and replaces the original Naive Bayes with the weighted attribute conditional probability,so that the classification results are as close to the actual category of the sample as possible.Compared with other Naive Bayesian classification methods,it shows that the proposed method has better classification performance for water quality data,and its accuracy,precision,recall and F1 value reach 96.0%,95.9%,93.8% and 94.8% respectively,which can be directly used in the water quality classification module of the water quality monitoring system.(2)Aiming at the problems of traditional water quality wireless monitoring technology,a lightweight water quality monitoring node based on NB-Io T is designed.First,the microcontroller completes the data acquisition of the water quality sensor through the Modbus protocol,and encapsulates a frame of the water quality data packet to be sent internally.Then,the terminal is equipped with the M5310-A module,which can use the relevant AT commands to send the data packet to the One NET cloud platform through the NB-Io T network.Through the interface of the One NET cloud platform,the real-time water quality data uploaded by the monitoring nodes can be viewed,and the read command can be actively issued to complete the data interaction.(3)Based on the mainstream framework integration of Spring + Spring MVC + My Batis,the construction of the water quality monitoring management platform has been completed.First,the receiving address of the water quality data is configured through the One NET cloud platform interface,so that the collected water quality data can be pushed to the application server through the Hyper Text Transfer Protocol(HTTP).Then,based on the push data of One NET cloud platform,the My SQL database is used to complete the data storage.Finally,a water quality monitoring management platform is built based on the water quality data reported by the monitoring nodes.Through the platform,users can not only intuitively view the water quality trend of each site,water quality classification results and its distribution in Baidu map,but also download water quality historical data reports,and actively issue reading command.