Vehicle classification using measurements from accelerometers mounted on the road surface

A new method for vehicle classification is presented. An axle parameter estimation gives advanced results for the distances between axles as well as transferred energy of vibrations. Measured road vibrations are analyzed and the phenomena are studied. Algorithms for vehicle classification based on the findings are described and their performance, advantages and disadvantages are evaluated and discussed. Correct classification of vehicles by wheelbases using accelerometers is about 89 %.