Numerical Simulation Of Urban Flood Based On Cellular Automata

Urban flood refers to the phenomenon of urban water accumulation caused by the large impermeable ground area and limited drainage capacity of drainage pipe network under the condition of continuous high-intensity rainfall.Due to the unusually severe global climate changes,such as the recent'El Nino',the rainfall capacity and intensity in some regions is constantly increasing,and in most of the cities in our country,factors such as serious aging,disrepair of underground drainage pipes and the low design standards result in urban surface water not being able to be discharged from drainage pipes in time under heavy rainfall conditions,which then results in urban waterlogging.In some cities,the drainage pipes are even the confluence of rain and sewage,and the pollutants will further reduce the flow area of the drainage pipes and even cause blockage.All these reasons increasingly make the city flood frequently happen,which then will always cause serious losses,and make the safety of people's lives and property under threat at all times.In order to study the formation mechanism and evolution process of urban flood,it is quite a necessity to carry out early warning,prediction and prevention of flood disaster,and to build the urban flood model for numerical analysis.However,due to a large number of mathematical and physical model equation parameters,as well as the complex calculation process and long simulation time,analytical solutions and more accurate simulation results cannot be obtained.This paper,based on cellular automata,a computer algorithm,constructs an alternative model to replace the mathematical and physical equation for the numerical simulation of urban flood,and then carries out an overall sensitivity analysis and calibration of many parameters of the urban flood model.In a word,this paper studies the actual characteristics and evolution rules of urban flood from another aspect,which provides reference for new methods and techniques ofurban flood simulation.The research results are as follows:1.The development status of urban flood models and cellular automata both at home and abroad is described in this paper,and the method of parameter overall sensitivity analysis and calibration is also introduced.2.With the method and principle of cellular automata and combining with the reality and characteristics of urban flood,the urban flood model,based on cellular automata,is constructed in this paper,and the mathematical programming software Matlab and Argic are introduced in detail to realize the model.3.This paper,taking the main urban area of Jinan city as an example,makes a sensitivity analysis of the model parameters by using two overall sensitive row analysis methods,namely the partial rank correlation method and mutual information method,and then conducts a comparative analysis to get the sensitivity of each parameter to the output result.4.The combination of automatic parameter calibration and manual calibration is realized.The artificial neural network is used to automatically calibrate the urban flood model based on cellular automata,and the automatic calibrating parameters are obtained.Then,according to the sensitivity analysis results of model parameters to perform manual calibration,after that,the model is verified with the measured data of main urban areas in Jinan,which shows the simulation results are perfectly in agreement with the measured data.5.The model is used to simulate the urban water accumulation area,the depth and time when heavy rainfall occurs in the main urban area of Jinan every 10 years,50 years and 100 years,which provides scientific guidance and basis for the early warning,prediction,emergency response and disaster prevention countermeasures of urban flood disaster.