Study On Analysis And Application Of Railway Dangerous Goods Transportation Risk Assessment Based On GANN

With the rapid development of China's economy, the needs of various industries on the dangerous goods increase year by year, the volume of railway dangerous goods is also growing year by year. many Intermediate links of the railway transportation of dangerous goods and high technical difficulty, and many types of dangerous goods, and complex operation conditions,make the risk of rail dangerous goods transportation high. Therefore, the comprehensive analysis of risk for railway dangerous goods transportation, quantitative understanding of the risks of railway dangerous goods transport system dangerous will be very significant.This paper first describes the research status of domestic and international rail dangerous goods transport risk assessment and situation of rail dangerous goods transport,on this basis, formation mechanism of the causes of transport accidents is analyzed.Then considering the risk factors of the railway dangerous goods transport and by-analysis of risk factors of the influence of the railway dangerous goods transport,with integration and optimization, this thesis establishs a more comprehensive, more targeted railway dangerous goods transport risk assessment system. According to the establishment of the railway transport of dangerous goods risk assessment indicators system,considering theoretical basis of the defects in the application of genetic algorithm and BP neural network, by the genetic algorithm to optimize the BP neural network, builds a more optimized genetic neural network model, which is efficient to achieve the rail transport of dangerous goods risk assessment. with the expert scoring method to construct the training sample, using MATLAB software tools to separately train the sample data for BP neural network and genetic neural network, the results of comparative analysis verify optimization of BP neural network by genetic algorithm feasibility. Finally this paper conducts risk assessment for the case through genetic neural network to verify the feasibility of the model,and executs sensitivity analysis for index parameters,to determine their influence degree to assessment results and provide direction to develop risk prevention measures.