Robot Execution Failure Prediction Using Incomplete Data

摘要：

Robust execution of robotic tasks is a difficult learning problem. Whereas correctly functioning sensors statements are consistent, partially corrupted or otherwise incomplete measurements will lead to inconsistencies within the robots learning model of the environment. So, methods of prediction (classification) of robot failure detection with erroneous or incomplete data deserve more attention. A probabilistic approach for the classification of incomplete data (which has three versions) is developed and evaluated using five robot execution failures datasets. We show that by improving the estimation of probabilities, our approach offers considerable computational savings and outperforms the other methods.

关键词： prediction incomplete data total probability theory decision tress robot failure detection

作者: Bhekisipho Twala

作者单位: Council for Scientific and Industrial Research (CSIR),P.O.Box 395,Pretoria 0001,South Africa

会议类型: 国际会议

会议名称: 2009 IEEE International Conference on Robotics and Biomimetics(2009 IEEE 机器人与仿生技术国际会议 ROBIO 2009)

会议地点: 桂林

会议语种:英文

页码: 1518-1523

在线出版日期: 2009-12-19（万方平台首次上网日期，不代表论文的发表时间）