Research And Design Of Household Odor Intelligent Monitoring System Based On Bionic Olfaction

With the rapid development of science and technology,people’s pursuit of comfortable and intelligent home life has also increased,and along with this increase,there are various indoor air pollution problems.Various harmful gases brought by uneven quality building materials and decorative materials,as well as the odors of cigarettes and inferior perfumes common in daily life,will lead to the continuous deterioration of indoor air quality and endanger the health of residents who live in it for a long time.The method of using a single sensor to monitor a single odor component in traditional household appliances is not enough to cope with the increasing indoor air pollution sources,and at the same time,it cannot meet people’s demand for an intelligent lifestyle.This topic relies on the existing smart home Io T cloud platform,combined with bionic smell and pattern recognition,to design a home smell intelligent monitoring system based on bionic smell.According to the types of household odors that need to be identified,an odor perception module that can be mounted on home appliances is constructed.At the same time,it proposes an odor monitoring solution that can be applied to smart home systems.Compared with the local machine identification of traditional home appliances,this solution performs the task of identifying home odors on the cloud platform,reducing the burden of operating home appliances.Compare the classification and recognition effect of BP-Adaboost method and traditional odor recognition method(PCA+LDA)on household odor.Finally,the system performance test was conducted.The specific work is as follows:1.Introduced and analyzed the domestic and foreign research status of indoor odor detection technology and bionic smell.2.Architecture design and smell monitoring scheme design of household odor intelligent monitoring system.3.The construction of the household abnormal odor characteristic database,including the design of the odor perception module as the collection end,for the particularity of the real environment,a feasible household abnormal odor sample collection scheme is proposed,and the final odor characteristic database is established.The establishment of the database provides a basis for the subsequent verification analysis of the odor recognition model.4.Comparative analysis of BP-Adaboost odor recognition method and traditional bionic olfactory odor recognition method(PCA+LDA),by comparing the recognition rate of the two to the test sample set,and comparing the BP-Adaboost classifier and its base classifier to the same batch of test samples Set recognition rate,to judge the feasibility of BP-Adaboost as a household odor recognition model.The results show that the recognition rate of BP-Adaboost for the home abnormal smell test sample set reaches 88.0%,which is higher than that of the test sample set using the PCA+LDA method or a single base classifier.Therefore,BP-Adaboost was chosen as the odor recognition model of the cloud platform.5.System performance testing,through testing the system functions and odor monitoring performance,the test results show that the household odor intelligent monitoring system based on bionic smell designed in this topic can realize the monitoring of common household abnormal odors.