A scaled down truck model equipped with a wireless dynamic charging system
was the main platform for development, in this study. The goal was to enhance
the efficiency of the power transfer from the road-way coil to the truck’s on-board
pick-up coil. This was done by following two methodologies; the first was to fine
tune the driving controls through an autonomous driving system, based on clas-
sic computer vision algorithms, in order to maintain a close to perfect alignment
between the two coils. The second methodology was to establish a real-time com-
munication link between the truck and the roadway controller; in order to syn-
chronize the triggering of the road-way coil to minimize the energy lost when the
truck is not passing over the road-way charger. The truck is equipped by an Nvidia
Jetson Tx2 board that handles the autonomous driving, and a Zynq board controls
the charging of the truck’s battery by rectifying and regulating the wirelessly re-
ceived power.On the sending side, the roadway unit is controlled by another Zynq
board, which was connected through bluetooth to the truck’s Jetson board. The
BT link enables the truck to trigger the charger on/off in real-time. Two different
triggering scenarios were tested; to quantify the parasitic idling losses and to il-
lustrate the utilization of the established real-time communication in minimizing
such losses. The development platform for the truck’s autopilot was ROS (robot
operating system), which is a powerful middleware framework that runs on Linux
based operating systems. As for the roadway unit, the Xilinx Vivado design suite
was used for development.