Behavioral rhythm conversion calculation and psychological crisis early warning based on environmental internet of things perception

Environmental internet of things can detect people's behavior through mobile phones or home sensors, providing a new way for early warning of psychological crisis. The existing methods are difficult to capture the change process of behavioral rhythm and rely on a large number of manually labeled data. This paper proposed a physics‑informed joint simulation‑recognition framework, which incorporates the biological clock law into the model to identify pre-crisis behavior patterns (e.g., increased nighttime activity and decreased socialization), and reduce label dependence through self-generated hard example training. The results show that the accuracy of the proposed method is 91.2% and 87.4%, which is about 8% and 6% higher than the best existing method, and the ability to correctly identify crisis individuals is increased by more than 10%. This provides a new idea for using internet of thing to achieve low-cost mental health monitoring.