Fundamental Experiment of Model Predictive Tracking Control by Hydraulic Excavator: (Suppression of Deviation from Target Shape Caused by Dislodging Unknown Object during Digging Operation)

K. Sekiguchi,Katsumasa Suzuki,Takumi Tomatsu,K. Nonaka
DOI: https://doi.org/10.5739/JFPS.49.1
Abstract:There are cases where a hydraulic excavator does not perform the accurate digging operation due to the influence of the resistance force when the bucket tip contacts with the unknown object in the ground. To resolve this problem, we propose a method combining model predictive control and servo mechanism. Model predictive control with servo mechanism, which we call as MPCS, predicts the future motion and suppresses the influence of the unknown disturbance. Thus, MPCS can remove the influence of the resistance force owing to servo mechanism which increases the control input using the error integration. Moreover, it suppresses the deviation from the target shape owing to model predictive control which predicts the future motion and considers constraints when the resistance force has disappeared. In this paper, we consider constraints of the magnitude of the control input and the maximum discharge flow rate of the pump in MPCS. We show the effectiveness of MPCS through the experiment including collision with the unknown objects.
Engineering,Computer Science
What problem does this paper attempt to address?