A Non-parametric Bayesian Network for multivariate probabilistic modelling of Weigh-in-Motion System Data

Weigh-in-motion (WIM) systems help to collect data such as vehicular loads, individual axle vehicle type, and number of axles. This is relevant in engineering because traffic load performs an essential function the design new bridges reliability assessment existing ones, analysis other areas engineering. Therefore, when WIM not available, computing synthetic observations that adequately approximate statistical dependence between variables important. In this paper, measurements from Netherlands Brazil were analysed, a set non-parametric Bayesian Networks (NPBNs) presented. paper significantly improves on previous results by allowing inter-axial distance be generated, several sources used modelling making software available researchers practitioners interested for generating based distribution type. particular, models describe weight length different types are derived. Three NPBNs quantified using from: (i) six locations motorway network Netherlands, (ii) one location city route Rotterdam, The (iii) highway Araranguá located south Brazil. Additionally, Graphical User Interface (GUI) Dutch motorways was developed. To illustrate possible use model available. GUI compute collected through counters gathered Toluca central Mexico, input. shows methodology here presented widely applicable depends only type configuration.