Safety Analysis Of Flash Floods Induced Landslide Disaster Relief Based On AHP-FCE

Flash flood disaster is a major type of natural disaster,with sudden and rapid development,large number of casualties,easy to secondary disasters,and difficult to rescue.In the process of disaster relief,due to the geological conditions of the disaster area usually caused the disaster area more blurred and complicated.Therefore,the rescue decision-making is essential,once the decision-making mistakes,it may cause rescue workers casualties,expand the degree of accident loss.Therefore,it is significant to study the safety dynamic evaluation method of the flash flood disaster rescue,which will provide the theoretical guidance for the rescue decision.This paper analyzes and summarizes the causes,influence factors,the types of disaster and spread characteristics of the flash flood disaster,and obtains the mechanism and rules of flash flood disaster.Based on the theory of modern accident causes,the safety evaluation index of flood disaster relief is carried out by using "man-machine-ring-management" system analysis method.The evaluation index system is optimized with 10 categories and 39 small class.Finally,based on the analysis and summary of the screening,optimization method of the evaluation index system,combined with the characteristics of evaluation target,using the Delphi method and AHP method to optimize the index system.At the same time,the method of index processing,quantization and weight assignment is put forward.According to the established evaluation index system,combined with AHP and fuzzy theory,AHP-FCE evaluation model is established;the model was tested by using the flood disaster relief of the bank of the Yangtze river in Badong county;Through the case data application,safety grade evaluation,provide scientific and accurate data for subsequent model test.Finally,the related concept and the knowledge of extreme learning machine were introduced,and the algorithm and the characteristics of several commom extreme learning machine were sumed up.According to the characteristics of flash flood disaster rescue safety evaluation and prediction,the algorithm I-ELM was chosen to be model based algorithm;then,the algorithm was improved and prediction model based on dynamic evaluation of improved I-ELM algorithm was established.Finally,using data from a sample of actual rescue case,has carried on the inspection to the established model,the scientificity and reliability of the model was verified.Through the research results of this paper,according to the disaster information acquired real-time in the process of rescue,the security of rescue personnel into the disaster was quantitative efficiently determined,which has great theoretical significance to commanding and decision-making of flash flood disaster rescue.