Abstract: The automatic online detection technology for household appliances abnormal sounds is being widely concerned by household appliances manufacturers. This paper presents a preliminary explanation of the technical approaches and system architecture for household appliances online automatic abnormal sound detection. The characteristics of the two technical approaches for the automatic detection of online abnormal sounds in household appliances are described and compared. The detected system architecture and the relationship between the various components of the architecture are briefly described. This article analyzes the key technical difficulties in the online automatic detection of appliance noise, and proposes a principled solution.

Key words: Household appliances abnormal sound, Machine learning, Online automatic detection