Resumen
The paper provides an overview of the development of a railway bridge weigh-in-motion (B-WIM) system, one of the first of its kind for weighing trains in motion. A steel truss bridge in Poland was used for testing the system. Four trains which passed over the bridge were weighed in a rail yard in Warsaw. The conventional road B-WIM system was adapted to calculate the weights of the train carriages using the measured response from the test bridge and the accuracy of the system was assessed. Initial result showed that weights of one of the four trains of known weight were predicted very accurately, but accuracy of the other three trains was poor, with calculated carriage weights deviating by as much as 30% from their actual values. An in-depth analysis showed that these trains were changing velocity as they traversed the bridge and that the large errors were directly correlated to this changing velocity. The standard B-WIM algorithm, which assumed a constant velocity during the passage of a vehicle or train, was adjusted to allow for the effect of this changing velocity. The results improved dramatically, with the vast majority of the calculated wagon weights falling within 5% of their actual values. Further developments tailored the B-WIM algorithm for weighing trains, including the system interface that employs graphics of locomotives and wagons. The development of the railway B-WIM has been a success and has demonstrated that calculations of train weights using instrumented bridges can be efficiently performed.