D-RANSAC: Distributed Robust Consensus
暂无分享,去创建一个
“Robustness is the ability of a system to cope with errors during the execution.” This property is essential in any robotic system. A reliable robotic network must be able to fuse its perception of the world in a robust way. Data association mistakes and measurement errors are some of the factors that can contribute to an incorrect consensus value. In this chapter, we present a distributed scheme for robust consensus in autonomous robotic networks. The method is inspired by the RANdom SAmple Consensus (RANSAC) algorithm. We study a distributed version of this algorithm that enables the robots to detect and discard the outlier observations during the computation of the consensus. The basic idea is to generate different hypotheses and vote for them using a dynamic consensus algorithm. Assuming that at least one hypothesis is initialized with only inliers, we show theoretically and with simulations that the studied method converges to the consensus of the inlier observations.