Highway access safety evaluation from macro-meso-micro levels

The existing of highway access is a critical factor that influences highway safety due to the interaction between access traffic from driveway and through traffic of arterial road. There are numerous factors that affect access safety such as road environment, land use, network accessibility, access density, access lane channelization, sight distance etc. This article presented an intergrated model to fully evaluate access safety from macro-, meso-and micro-levels. The resarch methodology had two steps. The first step: latent clustering analysis method was used to tease out key factors from numerous factors that affect access safety. The second step: the weight of key factors was obtained by structural equation modeling (SEM) that establish the relationship between highway access related crash date variables and key factors of access safety. Crash data provided by highway bureau of Jiangsu province, China was utilized for analysis. In total 3,483 access related crashes at national highway numbered G205 (occurring between 2008 and 2011) were considered in the SEM models. The proposed access safety evaluation model was then compared with the existing model by analyzing their correlation with crash rate on another national highway numbered G204 in Jiangsu province, China. The comparison result demonstrated that highway access related crash rate was more correlated with the proposed access safety evaluation model than that in the existing model. The proposed access safety evaluation model is promising in the usage of access design, renovation and management by highway administration and transportation consulting agencies.