Detection And Evaluation System Of Indoor Environmental Pollution Level Based On Integrated Learning

In contemporary society,indoor environmental pollution is one of the reasons of people’s health.Real-time monitoring of indoor environmental pollution has become a necessity in people’s lives nowadays.Formaldehyde is one of the major pollutants in the indoor environment and is extremely harmful.It can lead to respiratory tract infection,pneumonia,pulmonary edema,and even cancer.In recent years,there have been numerous health and safety problems caused by environmental pollution in indoor places.Therefore,this thesis designs a system to detect and evaluate the level of indoor environmental pollution.The system uses Wi-Fi networking to lay out environmental detection terminals indoors,enabling real-time indoor environmental detection and evaluation.The administrator takes use of the management system to manage multiple environmental detection terminals.The user views the pollution level of the current place on the APP platform.The system mainly contains four parts: environment detection terminal,server,APP client and indoor environment monitoring management system.The environmental detection terminal is used to detect the temperature and humidity in the indoor enviromnent,as well as pollution factors,such as formaldehyde,PM2.5,PM10 and ammonia.The detection data is sent to the server in real time through the wireless Wi-Fi module.The server receives the data and then uses the algorithm to proceed.Real-time evaluation,and stores the original data and evaluation results in the My SQL database.The APP client initiates a data acquisition request to the server by scanning the QR code of the environmental detection terminal to obtain the real-time detection data and evaluation results of the environmental detection terminal.The indoor environmental monitoring management system is used to manage the working status of the environmental detection terminal,view historical detection data,evaluation results and user information.The system uses Stacking integrated learning to combine fuzzy mathematical comprehensive evaluation algorithms with classification algorithms to obtain more accurate and objective evaluation results.The analysis shows that the fuzzy mathematical comprehensive evaluation algorithm is strongly subjective in the way it determines weights and affiliation functions,and cannot accurately distinguish the boundary data.However,the classical classification algorithm can obtain an objective classification result from the original data without considering the actual contextual meaning of the data.Therefore,using the Stacking integrated learning can combine the advantages of fuzzy mathematical integrated evaluation algorithms and classical classification algorithms,and overcome the disadvantages of the different algorithms.In this thesis,we use the Stacking integrated learning algorithm based on K-nearest neighbor to evaluate the pollution level by collecting data from different simulated environments.We verify the objectivity and accuracy of the evaluation method of this system by combining the actual data and the experimental environment.We conduct joint testing and verification of various parts of the system.The testing and verification results show that the system can accomplish the design objectives.