Ultrasonic Positioning System for Electric Road System

Abstract: Elonroad is a company that is looking into the possibility to charge electric vehicle on road. The conductor implemented on the driveway has a characteristic shape. The scope of this master thesis project is to identify the shape of the conductor and determine the relative position of the vehicle to the conductor using ultrasonic sensors. The goal is to have an update frequency of 30 Hz and a positioning error of maximum 2 cm. During the project a serial system containing 7 ultrasonic modules has been built. The sensor modules are mounted in an array and placed parallel to the road and perpendicular to the vehicles intended direction of travel. The ultrasonic modules are placed with a center distance of 10 cm on the array. A cross-correlation algorithm was implemented to determine the distance to the surface below each of the ultrasonic modules. The method to find the relative position of the vehicle uses a modified convolution algorithm which is proven to work under ideal circumstances. The distance measurement from the modules to ground can differ roughly +/-8 mm, this originates from the wavelength of the 40 kHz ultrasonic signal in air, the +/-8 mm error can, according to simulations give a positioning error of up to 5 cm. Simulations has indicated that +/-4 mm will give a positioning error less than, or equal to, 2 cm. The update frequency is assumed to be around 20 Hz, where the largest part is due to calculation time of the distance to ground. It can be lowered with more effective algorithms or with a more powerful microcontroller.