仿人机器人多传感器定位系统

为了提高仿人机器人SHFR-Ⅲ人机交互系统的定位精度和定位可靠性，设计多传感器定位系统，包括基于热释电红外传感器垂直阵列的红外定位系统、基于听觉传感器三角形阵列的听觉定位系统和基于立体双目视觉的视觉定位系统.提出权重随目标位置和外部环境可变的加权平均融合算法，将3种定位数据的融合结果作为最终定位结果.实验结果证明：多传感器定位系统的限制条件比单传感器定位系统更少，环境适应性更强，有效提高了交互系统的整体定位精度和可靠性.
Abstract: A multi-sensor localization system was designed in order to improve the positioning accuracy and reliability of human-computer interaction system for humanoid robot SHFR-Ⅲ. The system included an infrared positioning system based on vertical array of thermal pyroelectric infrared sensors, an aural positioning system based on triangular array of auditory sensors and a binocular vision positioning system. A weight fusion algorithm in which weights varied to adapt to target location and external environment was presented, and three sensors' data were fused as the final positioning result. The experimental results show that the multi-sensor positioning system has fewer restrictions and more environmental adaptability compared with a single-sensor positioning system. The system effectively improves the overall positioning accuracy and reliability of the interaction system.