Springe direkt zu Inhalt

Tom Bullmann:

Parkingbehaviour of RC-Cars - Comparison of A* and RRT


Driver assistance systems are common usage in newer vehicles and even autonomous cars are advancing further and further each year. Both are intended to help make car travel more secure. Be it warning the driver of getting too close to a car in front of him, giving a signal when crossing the side lane or parking automatically to avoid bumping into a parked car. Autonomous trucks are even driving on the highway in Nevada, USA.

One very essential yet basic feature of autonomous cars is determining the path to take to get from where they are to the target location, preferably without touching any obstacle in the way. This can be done using any one of numerous algorithms. I am going to compare two of those to determine an efficient and stable way to calculate paths without coming into contact with other obstacles.

Using a path finding algorithm for a basic maneuver as parking could be construed as overdoing, as parking is a relatively simple procedure. Yet it does bring advantages compared to a static procedure that executes the parking once the car gets into the right starting position. Probably the most important advantage is that the behaviour of the car is more flexible. All it needs to park is find an appropriate parking spot with enough space to maneuver in without touching any surroundings.

That way it does not matter what kind of parking space is being used. Whether it is parallel parking, horizontal or diagonal. As soon as the spot is found, the appropriate path will be calculated and the car parks without incident.

Bachelor of Science (B.Sc.)