With the rapid development of the Internet of Things technology, the application of smart home systems in modern families is becoming increasingly popular, which has greatly changed people’s lifestyle. However, while smart home systems have made significant progress in automation and functionality, existing systems still have many shortcomings in emotional interaction design. Users not only expect the smart home to have basic operating functions, but also expect the system to provide a more personalized interactive experience with emotional resonance through emotional recognition and emotional feedback. The theory of actor network provides an important analytical framework for this study. Through user interviews, questionnaires and behavior analysis, this study constructed an actor network model of emotional interaction in smart homes and identified the “forced crossing point”. The crossing point is a key node in emotional interaction design, covering the accurate capture of users’ emotional needs, seamless collaboration between devices, and the accuracy of AI emotion recognition. Based on the analysis of point, the research proposes an emotional interaction design strategy from the instinct level, the behavior level and the reflection level to optimize the sensory experience, interaction logic and continuous iteration. Future research can further refine the emotion interaction design strategy, and combine advanced technologies such as artificial intelligence and big data to achieve more accurate and personalized emotion recognition and feedback. At the same time, it is necessary to pay attention to the protection of user privacy and ensure that technological development does not harm the rights and interests of users.