Mission planning for a heterogeneous robotic system

U ovom radu proučen je način koordinacije više autonomnih robota s različitim sposobnostima koji zajedničkim snagama obavljaju misiju. Misija je reprezentirana TÆMS strukturom stabla, koja predstavlja individualne akcije robota i međuovisnosti akcija koje zajedno tvore misiju. Korišten je GPGP decentralizirani algoritam planiranja, gdje roboti zasebno planiraju svoje lokalno viđenje misije i koordinacijom s drugim robotima zajednički stvaraju plan izvođenja misije. Algoritmi su ispitani na više letjelica s različitim sposobnostima u simulacijskom okruženju Gazebo te je predložena metoda implementacije na stvarnom sustavu.This thesis discusses the method of coordinating multiple autonomous robots carrying out cooperative missions taking into account their unique capabilities. The mission is represented by the TÆMS structure that forms a tree-like structure of individual actions to be taken by the robots and their dependency. A GPGP decentralized planning algorithm was used, where robots independently plan their local mission view and coordinate with other robots jointly to create mission execution plans. The algorithms were tested on various drones with different capabilities in the simulation environment Gazebo and the implementation method on real systems has been proposed