The increasing integration of IoT sensors, fiber-optic sensing and wireless communication links into urban architecture has made smart city infrastructure more vulnerable to disruptions as these complex, dynamic operating environments grow increasingly diverse with connectivity. Traditional static management frameworks cannot solve the problem of real-time perception and comprehensive planning of cross-system hazards. This paper proposes a dynamic monitoring and emergency dispatch model based on high-precision spatio-temporal big data. A data twin view of an integrated system based on heterogeneous data in the form of IoT sensing, remote sensing, GIS, and social sensing is established with a common spatiotemporal reference point. Accordingly, multi-layer complex networks are used to describe infrastructure dependencies, and spatiotemporal graph convolutional networks are integrated to obtain dynamic risk propagation simulations. In addition, a better non-dominated sorting genetic algorithm is adopted for multi-objective adaptive emergency resource scheduling. The results show that the model’s anomaly identification accuracy is 95.3, risk propagation prediction accuracy is 88.7, and only the time to generate the schedule is 4.2 seconds. The proposed data fusion and scheduling logic is consistent with smart-city systems that rely on wireless sensing, remote sensing and communication infrastructure.
Zhang et al. (Thu,) studied this question.