机器人多传感器信息融合算法研究

A study of multi-sensor information fusion algorithm for robot

定义可信度区间估计，建立关联阵及可信度阵以检验传感器测量值的误差．采用有向图描述传感器之间的关系，确定传感器数据的取舍，提出两种最优融合算法，仿真结果表明优化算法是有效的，优越于简单平均法．

A confidence distance measure is defined,a relation matrix and a confidence measure matrix are proposed,in order to test errors of sensor measuring values. Using directed disgraph to describe the relations among sensors,whether datas of sensors are used are decided. Two Kinds of optimal algorithms are represented to get the optical fusion results. Simulation results have shown that the optimal algorithms are effective and are better than simple averaging method.