reinforcement learning parking system
This project uses uagame.py which is a module from the cmput 174 class at the University of Alberta.
Class Car: Drive: increases the velocity of the car by 1 unit (if the car is in forward motion): Reverse: decreases the velocity of the car by 1 unit (if the car is in backward motion): Right: turns the car wheels by 1 degree(unit) to the right: Left: turns the car wheels by 1 degree to the left: Stop: stops the movement of the car in any direction:
Optional fuel???
Object detection: front, back, left and right, tl, tr, br, bl
Car position is given by the centre of the car
If car moves into outside bounds it crashes and resets
If it moves into objects it also crashes and resets
Class map:
Made up of small grids which are each 1cm * 1cm The entire maps is 12 000 grids by 34 200 grids. Each parking spot is blah blah * blah blah The road size is 2 * parking spot size.