Parking problem for a differential drive robot

Naloga obravnava problem parkiranja mobilnega robota, pri katerem mora robot doseči določeno lego v prostoru. Zaznavati mora predmete oziroma ovire okoli sebe in najti primerno parkirno mesto za izvedbo parkiranja. Cilj je razviti program, ki mu bo omogočal samodejno parkiranje. Mobilni robot naj bo sposoben parkirati na več načinov: bočno, vzvratno, parkiranje na mestu in parkiranje med dvema pravokotnima stenama. Uporabljen je obstoječi mobilni robot z vgradnim računalnikom. Za opravljanje naloge si mobilni robot pomaga z infrardečimi senzorji in rotacijskima kodirnikoma. Program je napisan v programskem jeziku Python. Rezultati kažejo, da lahko mobilni robot uspešno opravlja več vrst parkiranja. Parkiranje na mestu in parkiranje pravokotno med dvema stenama mobilni robot izvede zanesljivo v vseh primerih. Pri bočnem in vzvratnem parkiranju pa se lahko pojavijo napake pri parkiranju, kot npr. nepopolna vzporednost mobilnega robota s steno. V splošnem pa z izdelanim programom mobilni robot parkira uspešno v 77,5 % primerov.The thesis deals with the parking problem in case of a differential-drive mobile robot. To perform a parking operation, the robot uses sensors to detect objects or obstacles in its environment, finds a suitable parking spot, and executes a parking maneuvre. The goal is to develop a program that will perform the parking operation autonomously. The mobile robot should be able to park in different ways: parallel, reverse, on-site parking, and parking between two rectangular walls. An existing mobile robot with built-in embedded computer is used. To accomplish the task, the mobile robot uses infrared sensors and rotary encoders. The program is written in Python programming language. The results show that the mobile robot can successfully perform several types of parking using the developed program. On- site parking and parking between two rectangular walls is reliably performed in all cases. However, in parallel and reverse parking, parking errors can occur, e.g. incomplete parallelism of the mobile robot end pose with respect to the wall. In general, with the developed program, the mobile robot parks successfully in 77,5 % of the cases