Indoor Environment Data Acquisition And Monitoring System Based On Internet Of Things And Machine Learning

With the rapid development of society,the concept of “green life”has been paid more and more attention recently.Along with the accumulation of all sorts of inhalable particles,fog haze and polluted gas from building materials,indoor environment is not as safe as people consider.At present,environmental problems have attracted wide attention in our daily life.However,neither can people obtain the environmental status data nor perceive the intractable problems.In view of this situation,an indoor environmental data acquisition and monitoring system was designed to obtain indoor environmental status data;and this system was based on the use of the development board and sensor devices.In the meantime,a mode of multi-user with different operation grades was designed through the visible presentation of the real-time/historical data via web sites.Besides,the stability of the sensor and the accuracy of the data were examined by the system.With regard to abnormal states of data,such as congestion,noise and abnormal fluctuation of data and return value anomaly,data state analysis was used to distinguish and lock the anomaly.With respect to the imbalance between sample sets of normal and abnormal data,the SMOTE algorithm is employed to balance the data,and SVM algorithm is used to classify and estimate the data.Simultaneously,corresponding models were trained and applied to the system,so as to enhance the safety and reliability of the system.In this thesis,the indoor environmental data acquisition and monitoring system is convenient for people to observe the indoor environment status visually and clearly.In addition,an economic and convenient way was utilized to provide a basis for indoor environment governance,narrowing the distance between the public and the environment.Therefore,the automation and universalization of indoor environmental monitoring can be resolved by adopting the Internet of things technology,Internet technology and machine learning techniques.