Flood Evacuation Migration Model And Application

Flood loss ranks the first of all kinds of natural disaster loss all over the world, so that the evacuation and path finding strategy in the flood disaster has been the focus of many scholars around the world. Route planning makes an important part of emergency refuge system, and it is widely used in the evacuation of hurricanes, fires, floods and other natural or man-made disasters. Existing relevant achievements at home and abroad to solve the practical problems encountered bottleneck. Under the circumstances that when the road network is too large or complex and the victims increasessharply, meanwhile, the possibility of the congestion in the process of evacuation and other specific cases there is no efficient solution.Besides, it suffers from high computational cost and may not scale to large transfer networks in practical evacuation scenarios. In this case, we propose two methods, and both of them can effectively solve the evacuation and route schedule problem in the flood protected zone or flood diversion area. Performance evaluation on the flood transfer network configurations shows that the model proposed in this paper produces high quality solutions, and significantly reduces the computational cost compared to linear programming algorithm. In comparison with the general linear programming algorithm, the latter proposed model saves computing resources and provides a strong basis and scientific support for decision-making in the flood transfer. The main work and accomplishments include:Analyzed the path planning strategy and key problems in the flood evacuation. Preparation for the proposed flood evacuation model included the decided transfer unit, available road and refuge area of the research area based on whether it has pre-arranged evacuation route schedule or not. Performance of the model's solution and the computational cost are the core part of the evacuation schedule, the models proposed in this paper provides the optimal solution based on comprehensive consideration of road grade, traffic congestion, limited refuge resources and so on.Network analysis and spatial analysis in GIS technology are widely used in flood evacuation plan research, but failed to scale up to large evacuation size, complex road networks, sharply increased victims and limited refuge resources in flood disaster scenarios. On account of this, this chapter put forward a flood evacuation migration model based on network flow, the victims and the refuge capacity are both being divided into specific consistent unit, which leads to the decreased computational dimensions and reduces the requirements on the performance of the computer, most of all, the model provides the optimal evacuation route with increased computation efficiency.Analyzed the influence of the divided unit in the flood evacuation migration model based on network flow that proposed in the third chapter on the results. Took the flood diversion and storage areas of Jing Jiang city as the research object, and route time-cost, route length, route flow were chosen to validate its effectiveness compared to the anterior evacuation scheme. The comparison shows that the new evacuation strategy reduces time-consuming, route length and relieves the congestion in the evacuation process which promote the transfer process. When the transfer unit defined to be 200, 300, 400, 500, the impact on the calculation result was not so obvious. So there is no need to divide the compute unit to a single person for sake of saving computation run-time and resources.Evacuation route planning tends to use linear programming approach, but the computation run-time and it's performance is sensitive and difficult to scale up to large evacuation networks. On the basis of a heuristic algorithm, namely Capacity Constrained Route Planner(CCRP), this paper put forward the incremental data structure, and considered the congestion delay in the intersections, which led to more precise simulation in the model construction of flood migration in reality. Besides, we compared the difference of whether there was computation priority in the proposed flood evacuation model. Achievements of the study have been applied to the national flood risk map project, and provide an important technical reference for other evacuation decision-making in disasters.