Abstract
Bridge weight-in-motion (BWIM) system is a method, that provides to identify axle weights. It is a non-destructive method, which allows not only to identify the axle weight, but it can show current shape of the structure, so it has a great potential. There are various methods to do measurements for this system. Mostly, accelerometers or strain gauge are used. Signal noise has significant effects to the results. It could be resonance of the bridge, wind, defect at the support system, defect at the roadway, etc. It is necessary to filter all this effects, to get clear data. There are many ways to do the filtering. Digital filters allow it. Sometimes, this type of filtering could remove important data about the crossing of the vehicle. It could generate inaccuracy of the whole system and create major errors to identified vehicles. It is necessary to find the optimal way, to keep important data and remove all dynamic noise. This paper will investigate the previously mentioned problems. Measurements will be accomplished on a small-scale model of the bridge. Vehicle will be crossing over the bridge, while the bridge will be awakened to the first vibration shape and other frequencies, that will have a great impact to the measurements.