Funding Reference LABELEC
Dates 2005-03|2005-11
Summary

Preventive detection of obstacle clearance infractions in power lines is important for electrical distribution and critical for transmission system operators. The project developed a prototype system to measure obstacle, in particular tree, clearance of high voltage electrical overhead lines based on laser scanning and performing on-line fault detection and classification.
The system features a video-camera, a GPS-based geo-reference and a low-cost laser scanner with a one-dimension angle sweep, all installed in a land and in aerial vehicle (helicopter). The motion of the vehicle while the laser scans sweeps a three-dimensional volume around the overhead line, generating a cloud of 3D points that represent, the line, the poles and the nearest object to the laser on any scanned direction. Experiments have been carried with four-wheel drive land vehicles and, most often, helicopters.
This solution relies on human-machine interaction for optimal results: the pilot steers the vehicle according to the inspector directions, adjusting speed, attitude and course, depending on the environment along the line track. The line inspectors role is empowered by accurate measurements and a clearance fault diagnosis system that support his/her assessment and focus the inspection on the critical spots, where he/she can acquire images and describe the scene by speech. The inspector must look for multiple faults along the line and the track. Thus, an optimized interface encompassing video, laser and grid layout is required not to hinder his performance.
The project developed an acquisition interface including laser, video and GPS data, a display interface, an algorithm to identify the line, the line supports and remaining obstacles from raw-data and a procedure to detect abnormal situations (e.g., obstacles closer to the line than a given threshold). Experiments were carried out the sensors on board an all-terrain vehicle and on-board the helicopter that performs the normal inspection operations for REN (National Electric Network).

Research Groups Intelligent Robots and Systems Group (IRSg)
Project Partners LABELEC
ISR/IST Responsible
Isabel Ribeiro