RoboCup サッカーロボットにおける局所的/大域的な位置推定を組み合わせた自己位置推定システムの開発

抄録

Various abilities are required for autonomous robots because they act judge and act on their own. The selflocalization problem that recognizes the surrounding environment and estimates the position is one of the most important tasks. The self-localization system is highly dependent on operating conditions, environment, and available sensors. That system is desirable operate stably for a long time. In this research, we focus on autonomous mobile robots for the RoboCup soccer meddle size league (MSL), but self-position kidnapping is a big problem. Therefore, we propose self-localization algorithm combining particle filter and perfect match, and aim for improvement of accuracy and stability of self-localization.