Full State Estimation for Helicopter Slung Load System

This paper presents the design of a state estimator system for a generic helicopter based slung load system. The estimator is designed to deliver full rigid body state information for both helicopter and load and is based on the unscented Kalman filter. Two different approaches are investigated: One based on a parameter free kinematic model and one based on a full aerodynamic helicopter and slung load model. The kinematic model approach uses acceleration and rate information from two Inertial Measurement Units, one on the helicopter and one on the load, to drive a simple kinematic model. A simple and effective virtual sensor method is developed to maintain the constraints imposed by the wires in the system. The full model based approach uses a complex aerodynamical model to describe the helicopter together with a generic rigid body model. This rigid body model is based on a redundant coordinate formulation and can be used to model all body to body slung load suspension systems. Both estimators include bias estimation for the accelerometers and gyros and the model based estimator furthermore includes estimation of external wind disturbances. A vision system is used to measure the motion of the load relative to the helicopter. A method is devised to reduce the execution time of the process model in the unscented Kalman filter. The two approaches are tested through simulation and compared. The full model based approach shows better results than the kinematic model aproach, but at the cost of a larger computational burden. I