| Urban floods are increasingly frequent due to the combined effects of climate change and rapid urbanization.Consequently,the mechanisms and patterns of disasters have also changed significantly,making flood control and prevention crucial for urban development,people’s safety,and social and economic development.As a result,urban flood control and prevention,along with advancements in information technology and intelligent emergency management,have become important topics for government officials and the academic community.Conducting research on flood risk avoidance path planning is an important step in improving disaster response capabilities,enhancing urban flood forecasting and warning systems,and promoting informatization and intelligence.However,the current research on risk avoidance path planning faces challenges due to the lack of flood information acquisition and other factors,which limit the availability of prior data for developing algorithms.This limitation significantly hampers progress in flood risk avoidance path planning.Fortunately,advancements in computer technology have made the 2D hydrodynamic model a valuable tool for flood forecasting and disaster management.This model incorporates strong physical process mechanisms and facilitates in-depth research into the urban rain-flood process.In order to address the problem of insufficient prior data in risk avoidance path planning,this paper focuses on utilizing an efficient and high-precision urban rain-flood numerical model.The research is conducted on the following topics:1 Establishing an efficient and high-precision urban flood numerical model:The analysis of production and confluence mechanisms in the urban rain-flood process,as well as the influence of fine topography on the confluence process,requires high-precision model input data to ensure simulation accuracy.This study introduces the use of advanced technologies such as Unmanned Aerial Vehicles(UAVs)equipped with laser radar technology,ortho/oblique photogrammetry techniques,and advanced rainfall and infiltration data measurement technologies to acquire high-precision digital elevation models.The construction of the urban rain-flood numerical model is discussed,covering aspects such as rainfall-runoff,surface runoff,numerical solution formats,and GPU acceleration technology.Additionally,the paper presents methods for assessing flood risks caused by unstable pedestrians and vehicle-induced factors.In conclusion,an efficient and high-resolution framework for constructing and simulating urban rain-flood models is established,incorporating high-resolution data collection,the urban rainflood numerical model,and risk assessment methods.This framework provides a theoretical and data foundation for research on risk avoidance planning.2 Constructing a grid-based pedestrian evacuation process simulation method:This research focuses on simulating pedestrian evacuation processes under flood disaster scenarios using a grid-based framework.The proposed framework combines the dynamic evolution process of flood disasters,cellular automata models,and artificial intelligence bee colony algorithms.A geographic cellular automata model is constructed to accurately capture flood dynamics and facilitate data storage and updates.By leveraging the strengths of swarm simulations in the artificial bee colony intelligent algorithm,the framework enables the simulation of pedestrian evacuation phenomena such as evacuation,instability,and detours during flood disasters.Two examples,the Shopping Centre and Senigallia,are used to simulate pedestrian evacuation processes with different initial population sizes.The efficiency of evacuation,instability rates,and other relevant factors are analyzed and evaluated.3 Constructing a road network-based risk avoidance path planning method:This study focuses on developing a path planning framework for pedestrians,vehicles,and other scenarios to ensure refuge during flood disasters.The proposed framework utilizes a road network-based approach that integrates numerical models of urban rainstorm processes,road network topology evolution,and shortest path planning methods.Additionally,a road network flood resistance assessment method based on a system function curve is introduced,considering variables such as effective node rate and effective road segment rate.By leveraging classic path planning algorithms like Dijkstra and Floyd algorithms,static refuge path planning(considering flood risk)and dynamic refuge path planning(considering the evolution of flood risk)are developed to facilitate algorithm implementation.4 Application research of refuge path planning for different scenarios of basin flooding and urban inundation:This study presents case studies involving the Morpeth flood disaster and the inundation disaster in Fengxi New City to investigate refuge path planning for pedestrians and rescue vehicles based on their respective flood disaster characteristics and the key objectives of flood disaster emergency management.Quantitative comparisons and analyses are conducted on risk assessment methods for pedestrian instability,as well as the computational efficiency differences between the Dijkstra and Floyd algorithms.Different flood disaster scenarios are established,and comprehensive analyses and comparisons are performed on indicators such as path planning distance and travel time.Furthermore,the computational time differences between static and dynamic path planning algorithms are assessed.The research findings aim to provide technical support for effective emergency management of flood disasters in specific local areas. |
