System Integration of Information Acquisition of Dynamic Environments and Caging based Cooperative Object Handling for Multiple Robot System

在本研究中，我们提出了一种基于笼型协同任务的信息采集系统，用于在动态环境下进行物体运输和搬运。为了表现人体的存在、姿态和运动历史，我们提出了一个混合多状态概率地图系统——人体运动地图的概念。针对多次实验获取的人体运动数据，提出了一种人体运动地图生成与高精度地图数据融合算法。实验结果验证了该系统的有效性和可扩展性。

In this research, we proposed an information acquisition system for performing caging based cooperative task, such as object transportation and handling in dynamic environments with human activates. For representing the human existence, posture and motion histories, we proposed a concept called Human Motion Map which is a hybrid multiple states probabilistic map system. A Human Motion Map generation and high accuracy map data fusion algorithm is developed for recording and integrating human motion data obtained via multiple experiments. Experiment results illustrate the validity and scalability of the proposed system.