|Inertial Parameter Estimation of an Excavator with Adaptive Updating Rule Using Performance Analysis of Kalman Filter
Kwang-seok Oh* and Ja-ho Seo
International Journal of Control, Automation, and Systems, vol. 16, no. 3, pp.1226-1238, 2018
Abstract : "This paper presents a rotational inertia estimation algorithm for excavators based on recursive leastsquares
with forgetting and an adaptive updating rule that uses the performance analysis of the Kalman filter.
Generally, excavators execute a swing motion with various materials, and the rotational inertia of the excavator
is changed greatly due to the excavator’s working posture. The large variation in the rotational inertia of the
excavator has an influence on the dynamic behaviors of the excavator, and an estimation of the excavator’s rotational
inertia is essential to developing a safety system based on prediction of dynamic behavior. Therefore, a real-time
rotational inertia estimation algorithm has been proposed in this study using a swing dynamic model. The proposed
estimation algorithm has been designed using only swing velocity, utilizing the recursive least squares method
with multiple forgetting for practical application to actual excavators. Two updating rules have been applied to
the estimation algorithm in order to enhance the estimation performance. The first proposed rule is the damping
coefficient updating rule. The second rule is the forgetting factor updating rule based on real-time analysis of linear
Kalman filter estimation performance. The performance evaluation of the estimation algorithm proposed in this
paper has been conducted based on the excavator’s typical dumping scenario. The performance evaluation results
show that the developed inertia estimation algorithm can estimate actual rotational inertia with the two designed
updating rules using only excavator swing velocity."
"Dumping scenario, forgetting factor, Kalman filter, recursive least squares (RLS), rotating inertia, updating rule."