Flood Evacuation Route Planning And Emergency Evacuation Planning Driven By Real-time Traffic Flow Information

Flood is one of the natural disasters which do harm to the lives and properties of people in our country.With the increasing of knowledge,the developing of forecasting technology and the innovation of scientific and technological means.we can get a deeper understanding of the nature of flood and therefore,can create a more accurate model to forecast flood.Nowadays,the study of flood control is prone to flood management instead of flood control.The measures for controlling flood have gradually changed from engineering measures to non-engineering measures.Disaster-proof system plays an important role in non-engineering flood control measures,which is the process of taking scientific evacuation scheme or emergency measures to reduce losses on the basis of flood warning.It is based on the evacuation route planning method and the emergency evacuation planning method.Flood control evacuation is a complex systematic project involving the large scale migration of thousands of people in a specific road traffic network.Not only the conflict between the macro instructions of the flood control management department and the driving choice of the masses themselves,the interaction among traffic entities in those complex road networks should also be considered.Moreover,it is also necessary to simulate the impact of traffic accidents on the network parameters of the evacuation road in the process of evacuation.Therefore,this is a highly complex nonlinear constrained optimization problem.And its modeling and solution has always been a hot while difficult problem in the field of disaster evacuation.With the evacuation route planning algorithm as the core,modern flood control evacuation needs to establish emergency evacuation planning and evacuation traffic management mechanism on the basis of accurate mastering of real-time traffic network information.Focusing on some key problems in evacuation route planning and the corresponding emergency evacuation planning,this paper establishes a system specifically to solve those problems,guided by intelligent calculation method,traffic assignment model and so on.The main contents and innovations of this paper are as follows:A traffic assignment model based on multi-scale fusion is established.Based on the cellular transmission model and car-following theory,a traffic allocation and simulation model is created which contains characteristics of multi-scale traffic model.The construction of traffic road network is realized by using network modeling language.In addition,a cell capacity model suitable for free flow traffic condition and a time cell model introducing the concept of semi-cell are proposed.And an improved vehicle clustering discrete packet unit model based on continuous packet motion equation is proposed.The vehicle-following equations using for simulating congestion wave and evacuation wave are established.The vehicle-following model referring to vehicle clustering unit is improved,which provides the underlying calculation framework for traffic allocation and simulation in flood-control affairs.An evacuation route planning algorithm based on specific quota allocation is proposed.Aiming at the complex topological structure of flood control evacuation network,based on the heuristic capacity constrained route planning algorithm,with different backgrounds as single source single sink,single source multi-sink and multi-source single sink evacuation scenarios which is based on the results of topology analysis,this paper established a multilevel objective optimization method with transit node buffer model,a quota allocation method which considers the length of time for emptying the road network and a confluenc quota allocation method based on particle swarm optimization.In addition,the current single shortest path search mechanism is extended from spatio temporal dimension to optimal path candidate set search,and is integrated into a new evacuation route planning algorithm CQRP,which can output high quality transfer scheme.Those are the core calculation methods for flood control evacuation route planning and emergency evacuation route planning.A fast real-time traffic flow information acquisition method based on video analysis is proposed.To meet the demand of real-time traffic flow data using in emergency evacuation planning and traffic management for evacuation,a real-time traffic flow information acquisition system based on video analysis is created.In analyzing the demand of real-time traffic network information for evacuation planning,a fast background update mechanism of partition block based on new statistical marker is proposed,which greatly reduced redundant computation in background update.The ghost problem of background updating of sequence images is analyzed from the aspect of mathematical model and a ghost suppression mechanism based on window analyzer is proposed,which increase the authenticity of the background images.Aiming at the rain and fog weather which usually happens in the situation of flood control and evacuation,and causes wet roads reflection,a reflection suppression algorithm based on continuous frame profile analysis is developed.By introducing CUDA acceleration technology and based on the dark channel de-fogging model,an image de-fogging algorithm which can adapt natural light automobile is proposed,which enables the system to collect the traffic flow information accurately in rainy days and foggy days.To meet the demand of real-time and accuracy of traffic flow statistics data,an instantaneous congestion detection method based on high-order frame difference function is proposed.In addition,we create an auto-transfer detection line and a dual-mode vehicle counting mechanism of virtual coil which can work when congestion happens.This greatly improves the efficiency both in vehicle matching and in vehicle counting.In this paper,the emergency evacuation planning method and the traffic management mechanism for evacuation are proposed.We create the method of simulating traffic emergencies and establish the fusion mechanism of multi-source traffic information data which can assist the decision-making for emergency evacuation.The traffic incident reporting mechanism based on intelligent equipment is proposed.Based on CQRP,the optimization principle and calculation process in the task of emergency evacuation planning are designed.At the same time,in cross roads,a human-froced traffic management unit mechanism driven by real-time traffic flow data is established to ensure the evacuation planning working orderly.The effectiveness of the algorithms proposed in this paper is verified.By simulating the calculation of the two-vehicle unit,the accuracy and validity of the vehicle clustering model proposed in this paper are verified.Through the simulation of congestion wave and evacuation wave in the time of cutting off the road,the effectiveness of the improved IDM vehicle following model proposed in this paper is verified.Through the experiment of collecting the traffic monitoring videos of multi-scene and analysis of the corresponding results,the effectiveness of the traffic flow information collection system proposed in this paper is verified.Through the calculation of the simulated evacuation scene and the calculation of the emergency evacuation scene in micro road network,the effectiveness of the CQRP evacuation route planning algorithm and the emergency evacuation planning algorithm proposed in this paper is verified.