This paper aims to develop a model to improve road safety based on the spatial and temporal features of accident occurrence. The spatial distribution characteristic of an accident follows a probability density function (PDF) that can be estimated with parametric functions. Estimating the PDF of accidents can help road specialists recognise accident-prone locations and improve road safety through countermeasures. In this study, the PDF of accidents is estimated using Gaussian mixture models. Gaussian models possess local and density information about accident distributions. For the prediction step, the parameters of the Gaussian models are predicted by means of a recursive least-squares approach. The major feature of this study is that the proposed model can accurately predict physical points where accidents are more likely to take place in the upcoming year. Moreover, because of the non-stationary feature of accidents, the proposed model is online, implying that it would update itself annually. The range of errors obtained from the results of cross-validation with existing data provides a very good indication of the accuracy of the proposed model.

No abstract

No abstract

No abstract

No abstract