Biosensor Technology Enables Intelligent Monitoring of Tourist Scenic Area Environment and Immersive Experience

Fu Maozheng
Hainan Vocational University of Science and Technology, Haikou, 571126, China
Zhang Jiani
Hainan Vocational University of Science and Technology, Haikou, 571126, China.
Zhang Yi
Hainan Vocational University of Science and Technology, Haikou, 571126, China.

Abstract:

This paper aims to solve the problem of personalized route recommendation in the immersive experience of tourist scenic areas, build an intelligent monitoring system based on biosensor technology, collect tourists' heart rate and other physiological data through wearable biosensor equipment, and combine environmental data such as scenic area temperature and humidity and tourist flow to construct multimodal feature input. STGAT (Spatial-Temporal Graph Attention Network) is used for spatiotemporal modeling to predict tourists' interest changes and comfort scores, and a reinforcement learning strategy is trained based on DDPG (Deep Deterministic Policy Gradient) to adjust the recommended path dynamically. Experiments show that the accuracy of tourist interest prediction of this method reaches 85.2% when the time window is 5 minutes, and the RMSE (Root Mean Squared Error) of comfort prediction reaches 0.45 when the time window is 5 minutes. In addition, the balance of tourist flow in the recommended path has been significantly improved compared to traditional Dijkstra and A * algorithms, and the distribution of tourist numbers at each attraction is relatively even. The recommended methods for this paper, Dijkstra, and A * attractions have an average stay time of 14.2 minutes, 12.4 minutes, and 13.0 minutes. This method can more effectively attract tourists to stop and improve their immersive experience, optimize scenic area management, and reduce the overload phenomenon of hot spots.