Fault Detection System And Diagnostic Model Of The Intelligent Home System Based On Big Data

Intelligent home system,combination of components has become irreplaceable in people's life.Its operating state could have a direct impact on people.Once the fault occurs,it could seriously damage the life experience,and even cause casualties in severe cases.At present,the detection of intelligent devices is still based on artificial recognition,and there is little research on fault warning and diagnosis system of intelligent devices.Therefore,it is of great significance to study the detection and diagnosis system with good correspondence,and to take timely treatment of household faults.In this paper,an intelligent washing machine is used as an example to study the fault diagnosis technology of inverter and asynchronous motor.The research contents are followed.Firstly,the main circuit of the inverter and the basic structure and operation principle of the induction motor were analyzed.The causes of various failures were classified.The diagnostic model was determined.For the large amount of data needed,in addition to obtaining from the database and establishing the fault simulation model,it was also collected by laboratory means.One part for training and one part for used,for comparison purposes..Then the three-phase output characteristics of the main circuit of inverter and the stator current characteristics of asynchronous motor are processed by using wavelet packet analysis method.The sample is collected as fault information.The feature weight of fault signal was extracted by wavelet packet analysis,and the feature vector was constructed.Finally,the characteristics and principles of GBDT,XGboost and SVM were analyzed.The parameters are adjusted by the method of grid search traversal.They were trained with data from the lab UCI database.The fault diagnosis of inverter main circuit,asynchronous motor and inverter-motor was carried out with the feature vector obtained by wavelet packet analysis as input and the fault results of inverter and asynchronous motor as output.XGBoost performs well against intelligent washing machines.In this paper,the circuit simulation on Matlab,wavelet packet analysis,SVM,decision tree applications to the diagnosis of home appliance fault,to solve the problem that difficult to accurately and efficiently diagnose when the fault occurs.The fault diagnosis of intelligent home system has been practiced to some extent.The feasibility of fault diagnosis for the whole system of intelligent home is confirmed.