Robot sürüleri için konum noktası tabanlı toplanma yöntemi

Aggregation, a widely observed behavior in social insects, is the gathering of individuals at any location or on a cue. The former being called self-organized aggregation, and the latter being called cue-based aggregation. One of the fascinating examples of cue-based aggregation is the thermotactic behavior of young honeybees. Young honeybees aggregate on optimal temperature zones in the hive using a simple set of behaviors. The state-of-the-art cue-based aggregation method BEECLUST was derived based on these behaviors. The BEECLUST method is a very simple yet very capable method with favorable characteristics such as robustness to noise and simplicity. However, the BEECLUST method does not perform well in low robot population densities. In this thesis, inspired by the navigation techniques used by ants and bees, a self-adaptive landmark-based aggregation method is proposed. In this method, robots use landmarks in the environment to locate the cue once they "learn" the relative position of the cue with respect to the landmark. Robots were utilized with odometry sensors to make the calculation of traveled distances possible. With the introduction of an error threshold parameter, the method also becomes adaptive to changes in the environment. In order to make robots robust to sensor noises and free of fine-tuning, reinforcement learning algorithm was employed to aid robots in coping better with uncertainties. In order to solve exploration-exploitation dilemma in reinforcement learning, a new cyclical update schedule was proposed. Through systematic experiments in kinematic and realistic simulators and real swarm robots with different parameters, it was observed that using the information of the landmarks makes the proposed method outperform other state-of-the-art cue-based aggregation methods such as BEECLUST and ODOCLUST in all the settings. It was also shown that utilizing reinforcement learning in the proposed aggregation method had a 20% performance increase in non-stationary environments. Additionally, reinforcement learning made the proposed method more robust to odometry noise reaching up to 30% performance increase.