Finding Bearing in Robot Navigation with the Use of the Kalman Filter

Abstract

In robotic systems navigation, it is necessary to determine the direction, in which the local coordinate system of a robot is to be turned in regards to the reference, global system. That direction from mathematical point of view is the rotation along one axis of the reference system, such that the reference system coincides with the local system associated with the body, which direction is defined. There is an assumption that the rotations along two other axes do not exist or they were taken into account in the calculations. A question appears: what is the measurement of the direction based on? The answer to that question is provided by nature [1, , because it has already solved this problem long ago. In migratory species such as birds, bees and fish, evolved magnetoreception abilities which allow an animal to detect the magnetic field of Earth. These abilities enable such organisms to navigate in space. In many devices magnetic field sensors called magnetometers are already used. However, magnetic measurements are subject to many kinds of distortions and errors. This paper shows a practical approach to a robot navigation problem.