Spatio-temporal awareness data, which is characterized by its spatial attributes, temporal attributes, and perception characteristics, has become increasingly important with the advancement of information technology, intelligent Internet of Things (IoT) technology, and machine learning theory. Spatio-temporal awareness data offers a new perspective for emergency disaster rescue, as it provides insights into the location, timing, and sensory information of various events and occurrences. By leveraging Spatio-temporal awareness data, emergency responders can better understand the nature and extent of a disaster, and make more informed decisions about how to respond to it. The integration of Spatio-temporal awareness data into emergency disaster rescue efforts has the potential to significantly improve the efficiency and effectiveness of these operations. The comprehensive use of Spatio-temporal awareness data for situational information sensing and extraction in disaster areas can form a new focus point for emergency rescue. In the emergency rescue problem, the scientific emergency evacuation of the affected population is the key to the rescue problem. As Spatio-temporal awareness data has rich semantic information and high practical value, and has a positive guiding effect on disaster rescue, it is a practical task to study and use Spatio-temporal awareness data for emergency evacuation. This paper focuses on the problem of resource-constrained crowd emergency evacuation planning under Spatio-temporal awareness data, and investigates the model and method of evacuation site confirmation and crowd emergency evacuation planning based on Spatio-temporal awareness data, with a new view to providing positive ideas and technical assurance for emergency rescue. Firstly, the role of Spatio-temporal awareness data in emergency crowd evacuation is analyzed, which can assist emergency decision-making. Secondly, an emergency evacuation framework using a rolling time-domain planning strategy based on Spatio-temporal awareness data is proposed, and an emergency evacuation place recommendation algorithm based on pedestrian flow prediction is designed. An integer programming model of emergency evacuation population allocation considering evacuation preference is also constructed to meet the satisfaction of different evacuation populations with evacuation places during emergency evacuation. Then, a global optimal evacuation path planning algorithm is proposed, and the emergency evacuation path planning problem is transformed into a multi-group evacuation path planning query problem, and a solution algorithm for the global optimal evacuation path planning query is further designed to solve the multi-group evacuation path planning query problem by using a network expansion strategy and a pruning strategy. In addition, in order to improve the quality of the emergency evacuation path search results, an improved algorithm based on refinement operation is also designed to improve and optimize the path combinations generated by the global optimal evacuation path planning query algorithm. Finally, through extensive experimental analysis and evaluation, the effectiveness and practicability of the proposed method and model are fully verified, which can significantly shorten the global travel time. In conclusion, this paper mainly studies the overall technical streamline of the emergency evacuation process, including proposing the evacuation site confirmation method and model, designing the crowd distribution model and emergency evacuation path planning method, in order to provide necessary technical support for emergency evacuation management. © 2023 Science Press. All rights reserved.
