Empirically-based modelling approaches to the truck weigh-in-motion problem

The paper develops and compares a comprehensive range of configurations of empirical modeling techniques for solving the truck classification by weigh-in-motion problem. A review of existing artificial neural network approaches to the problem is followed by an in-depth comparison with support vector machines. Three main model formats are considered: (i) a monolithic structure with a one versus all strategy for selecting truck type; (ii) an array of sub-models each dedicated to one truck type with a one versus all truck type selection strategy; and (iii) an array of sub-models each dedicated to selecting between pairs of trucks. Overall, the SVM approach was found to outperform the ANN based models. The paper concludes with some suggestions for extending the work to a broader scope of problems.