Collective error detection of onboard intelligent compasses by consensus agreement algorithm

In this paper, a new frameworks for error detection of intelligence compasses which are installed on multiple robots by consensus agreement algorithm is proposed. Intelligent compass´s accuracy is sometimes too unsteady to detect its fault. Therefore new error detection method is a major requirements for its long term application and coordination of multiple robots. This new error detection method is based on cooperative behavior of multiple robots. If sufficient number of robots head in the same direction, it is easy to detect wrong sensors. The alignment of robots by hand is too a heavy physical labor to operate a large number of robots. The proposed adopts average consensus algorithm for its synchronization so that no human assistance is required. Also no individual identification capability of robot for this method is required which gives it high scalability. After actual data of 3 intelligent compasses are shown, our framework which suggests effects of noisy input by local observation of each robot is proposed. Finally, by a series of computer simulations the hopeful characteristics of the proposed is confirmed.