Voice interaction-driven smart home fault diagnosis system for voice-driven smart home diagnostics

In order to solve the problems of insufficient intelligence and poor user experience in the process of smart home fault diagnosis, this paper proposes a voice interaction-driven smart home fault diagnosis system. The system first uses voice interaction technology to understand the user's fault diagnosis needs, and then constructs a home fault knowledge graph containing multimodal information such as home information, equipment fault information, processing methods and maintenance records. Through this knowledge graph, the system can accurately locate the cause of the fault and feedback the fault solution to the user in the form of voice interaction. The experimental results show that the FunASR voice recognition model and the fault diagnosis method based on the knowledge graph are the best.