The ability to know the location (localization) of robots in the environment or other robots, play an important role in multi-robot systems. Many methods and sensors had been propose, for example Dead-Reckoning; GPS; Vision based; etc.. Dead-Reckoning produce an accumulation errors, if used in a wide area or in a long time moving[1]. GPS has a level of significant errors when applied to a small robot or on the environment that are not large [2], while the vision has a poor response when applied to a fast-moving robots [3]. This research applied the trilaterasi methods that have been commonly used in communications systems, for localization in a multi-robot system composed of robots +11 cm in diameter. Sensors that are used in the form of 3 receiver with relatively fixed position to each other on a robot and an ultrasonic transmitter on the other robots. The principle of measurement based on the arrival time difference, or TDOA and processed by equation trilaterasi [4], to get the distance and relative orientation of the robot. Tests conducted on the condition still and moving the robot, at a certain distance and orientation, with a range of distance measurements between 30cm to 180cm. The result shows that this method successfully applied to the robot, with the average accuracy of 6% in the room, and the average accuracy of 5,2% outside the room. From the results of the analysis is also known that due to lack of the presence of some obstacles, that generate a magnetic field and the significant temperature changes in the robot environment. Another weakness of the system this system can not be used the same sensors (ultrasonic) for the purposes of obstacle detection, because it causes the interference.
展开▼
