I recently got the request to determine contact wire heights from multiple years of point clouds with the goal of doing a trend analysis. Point clouds acquired in four years were available, all of them acquired with a train-mounted mobile mapping system. The computation of the contact wire height is automated en consists of the following steps:
- Merging the point clouds of the two laser scanners and cutting them into 100m long sections.
- Classification of the point clouds.
- Modelling of rails and computation of track centerline and cant.
- Modelling of contact wire.
- Computation of eccentricity and height of the contact wire with regard to the track centerline.
The resulting plot clearly shows that the contact wire height increases throughout the years. This is probably the result of subsidence of the track.
The advantage of these surveys compared to a survey train that measures the wire height through the pantograph is that the contact wire is not pushed upwards. Additionally, it is possible to model wire that are several meters distant from the driven track. This not only allows for modelling track and wires of neighbouring tracks from a single ride, it is also possible to track model incoming and outgoing wires at tensioner locations (see 13200-13400m). This allows for checking whether the incoming wire is lower than the contact wire, which is a serious calamity risk.
