Unscented Kalman Filter for Real-Time Load Swing Estimation of Container Cranes Using Rope Forces

Abstract

The container crane represents the link between the containership and the port. It dictates the general conditions for the efficiency of container handling from the ship to the land and vice versa. While containers are handled by the crane, load swing reduces the rate of container turnover. In order to reduce load swing control systems are employed. Closed-loop control systems contain devices to track the position of the load with respect to the trolley's position. Accurate tracking of the load's motion during operation requires additionally installed sensors. Alternatively, the principle of state estimation can be employed. The observation of the motion of the container is carried out by a system model in parallel to the real system, taking into account the available rope force sensor information. Both, nonlinear system model and nonlinear sensor model are taken into consideration. An unscented Kalman filter is designed to estimate the states of the motion of the load. The observer is validated at the container crane test stand in order to provide accurate states for load swing control. Results are presented and discussed.