Study On Risk Assessment Methods Of Freeway

Risk assessment of freeway is one of the most important aspects in traffic safetymanagement field. At present, the most common risk assessment methods of freeway areanalytic hierarchy process, gray measuring methods, fuzzy comprehensive evaluation method.With the development of several years, these risk assessment methods are relatively mature,but compared to precise assessment result, these methods did not do well. Based on thosefacts, new risk assessment method will be present while computer engineering andmathematics are developing rapidly. New risk assessment method can use basic data to giveus a better assessment result, so it provide us a new way to solve the problem of traffic safetyfield.This article analyzed three critical safety factors from the point of road system, thesemain safety factors are people, vehicle, road, traffic environment. This article also elaboratedhow the branch factors of main factors like speeding, fatigue driving, automotive dynamics,vehicle stability, number of adverse weather, traffic flow affect safety of freeway and whatrisk these branch factors will bring to freeway.This article introduced basic theory of the most common risk assessment methods offreeway in China, like analytic hierarchy process, gray measuring methods, fuzzycomprehensive evaluation method, and compared with each other. Based on thoseintroductions, this paper proposed a new risk assessment method, radial basis function neuralnetwork assessment method, and then compared new assessment method with old method, inconclusion, new assessment method is more advanced than old assessment method, at last,this paper elaborate its mathematics theory and how to use new assessment method in reality.Using the data of Lan Ding freeway which is a part of Lian Huo freeway start form Lianyungang to Huoerguosi, this paper assessed Lan Ding freeway by radial basis function neuralnetwork assessment method. Established assessment indicator system by traffic environment,traffic flow, and other data of Lan Ding freeway, and then nondimensionalized theseassessment criteria, at last established mathematical model by scientific software andcalculated assessment result. Assessment process has showed that in order to improve accuracy of risk assessment offreeway, new method of risk assessment must be present, much more scientific and accuratemeans will be used to assess risk level of freeway. Radial basis function neural networkassessment method will make the assessment result more close to the truth and can tell uswhether the evaluated freeway system is safe, and also provide the basic information forimproving freeway system...