Real-Time Parameter Estimation for Arbitrary Aerial Manipulators

In this paper we present a novel algorithm for estimating system parameters of arbitrary aerial manipulators. We abstract an aerial manipulator as a rigid body with forces and torques applied at arbitrary locations. It is also assumed that there are sensors that measure forces, accelerations and angular velocities at these locations. The problem studied here is to fuse these sensor data in real time to estimate mass, location of center of gravity, moment of inertia matrix and the locations where these forces are applied. We analytically derive expressions that relate these quantities to the sensor data instantly or over consecutive samples. We also characterize the motion of the manipulator for which such estimates are impossible to obtain. This characterization allows us to define maneuvers for the aerial manipulator for which we can always determine the rigid body parameters. The algorithm is successfully demonstrated on a quad-rotor system with a moving arm, which causes real-time variation in the inertia matrix and the center-of-gravity location.