| With the rapid development of the national economy and increase of the people’s materialdemand, more and more dangerous chemicals enter into people’s life, Which makes thecirculation of the dangerous goods increased year by year. Compared to road transportation,railway transportation and water transportation has too much restrictions, therefore, thevolume of dangerous goods traffic by road transportation is increasing year by year. However,due to the high randomness exist in the process of road transportation, complex types of thedangerous goods and most of the transportation needs across the land, result in the roadtransportation has large risk. So, quantitative research the Risk factor on road transportationsystem of dangerous is of great significance.Factors that affect on the safety of dangerous goods transportation include two aspects:one is the packing of the dangerous goods; the other one is the external environment.Packaging of dangerous goods has the function of protection within the product andconvenient storage and transportation of the products, however, once the packing ofdangerous goods is damaged In the process of transportation, that often lead to transportaccidents. There are two reasons often lead to package in damaged condition of dangerousgoods: one is Packaging itself reason, such as sealing, design and so on; another is because ofthe impact of external factors, such as temperature, humidity and so on.This article is based on the topology structure and characteristics of BP neural network, byfocusing on analysis the influence of external factors to the dangerous goods package in theprocess of road transport action, and then analysis the influence of the external factors to roadtransportation of dangerous goods safety. At last, Construct a risk evaluation system aboutroad transportation safety of dangerous goods, which mainly On the basis of transportpackaging, and with the aid of MATLAB simulation software, to verify the feasibility andpracticability of the risk evaluation system.Call the trained network on risk evaluation ofexamples, and then tested the sensitivity of the output to determine the influence degree ofeach index on the output results, to provide a theoretical basis for risk prevention. |