Research And Implementation Of Mountain Flood Warning In Improved SCEM-UA Algorithm And Xin’anjiang Model

In this paper, with improved SCEM-UA algorithm and Xin’anjiang model are applied, developing mountain torrent disaster warning system based on the improved SCEM-UA algorithm and Xin’anjiang model in Chongqing Banan.The research work of this paper has done is as follows:(1)It introduces the concept of hydrological models, the structure and model application technology; It expounds the structure, the calculation principle and method of Xin’anjiang model, parameter and treatment of the model rate, production flow and evaporation, water division and bus processing technology; It discusses the theoretical basis of calculus Masjenet model, calculus equation and continuous algorithm, methods of calculating principle and derive the unit of analysis.(2) The basic concept of Shuffled Complex Evolution-University of Arizona algorithm,algorithm flow, and Markov Chain Monte Carlo method of bayesian inference are analyzed, markov chain theory and several kinds of commonly uses Markov Chain Monte Carlo sampling method; Based on the above theory, the Shuffled Complex Evolution Metropolis-University of Arizona algorithm algorithm is improved. The experimental results show that the improved Shuffled Complex Evolution Metropolis-University of Arizona algorithm performance is better than traditional Shuffled Complex Evolution Metropolis-University of Arizona algorithm algorithm, improve the efficiency of the algorithm.(3) SCEM-UA algorithm and the application of Xin’anjiang model in banan runoff estimating is researched. The experimental results show that improved SCEM-UA algorithm is superior to the traditional SCEM-UA algorithm in convergence efficiency objective,function and stability; the calculated runoff of Xin’anjiang model and the actual measured runoff has good performance in the match.(4)With developing and verifing,new early warning model is suitable for the wubu river valley in banan, and implements the Chongqing banan mountain torrent disaster warning system. With using the good of middleware technology in the design and using GIS map service, the design efficiency and visibility of the system is improved.