Optimal Excavation considering Operation Progress using Model Predictive Control for Hydraulic Excavators

Automatic digging operation utilizing information of surrounding environments for hydraulic excavators has been studied. Automatic digging is realized as a tracking control where the desirable path is given in advance. In addition, approaches based on optimal calculation for designing the digging path have been proposed. When the amount of remaining soil is larger than the capacity of a bucket, multiple excavations are necessary: a method to design an optimal digging path corresponding to each digging motion is required. This paper proposes a method that achieves the target shape gradually and automatically by considering remaining soil. Focusing on that the remaining soil decreases by each digging, the soil amount is utilized as a weight for index function. By using the evaluation reflecting decrease of the remaining soil, the digging path approaches the target shape at every excavation. The effectiveness of the proposed method is verified through simulations and experiments with a hydraulic excavator.