Load Position Estimation for Crane Anti-Sway Control Systems

Abstract

A rising number of modern cranes are equipped with anti-sway control systems to facilitate crane operation, improve positioning accuracy, and increase turnover. Commonly, these industrial crane control systems require pendulum state information for feedback control. Therefore, a pendulum sway sensor (e.g., a rope-mounted gyroscope) and a signal processing algorithm are required. Such a signal processing algorithm needs to filter out disturbances from both the sensor and the crane, e.g., signal noise and string oscillations of a long rope. Typically, these signal processing algorithms require the knowledge of the acceleration of the rope suspension point. This acceleration signal is often estimated from drive models. When drive models are uncertain, the pendulum state estimation accuracy suffers from drive model inaccuracy. In this contribution, an improved estimation algorithm is presented which estimates the load position without relying on the rope suspension point acceleration. The developed Extended Kalman Filter is implemented on a Liebherr mobile harbor crane and its effectiveness is validated with multiple test rides and GPS load position reference measurements.