Online payload estimation for hydraulically actuated manipulators

Abstract

Hydraulically actuated manipulators are commonly used for material handling or at construction sites due to the high power density of hydraulic systems. Dig assistance systems or payload monitoring systems for excavators become more common to increase the efficiency and productivity. In this paper, a novel approach for estimating the dynamic parameters of a payload (including the mass) attached to the last link of a hydraulically actuated rigid body manipulator is proposed. Motion equations for the open chain rigid body system including the payload are derived. The motion equations are formulated as a linear regression model exploiting the linearity in the base dynamic parameters. Furthermore, a dynamic model for the neglected closed kinematic chains and the friction is proposed. A least squares optimization problem for minimizing the error between the model based torque and measured torque with respect to the payload parameters is defined. The optimization problem is simplified with assumptions on the center of gravity and the inertia of the payload, which leads to a more robust estimation of the payload mass with a recursive least squares algorithm. Post processing of the estimated payload is suggested and heuristics for the calculation of characteristic values like the accumulated handled mass are shown for the special case of a material handling and an earthmoving excavator. Finally, the novel approach for the payload estimation is validated on both experimental excavators for typical working cycles. The overall performance meets the requirements of ± 3%.