Title: Event-based detection of the digging operation states of a wheel loader earth moving equipment

Authors: Salem A. Haggag; Nabila S. Elnahas

Addresses: Ain Shams University, Department of Automotive Engineering, 1 El-Sarayat St., Abbassia, Cairo 11517, Egypt ' Ain Shams University, Department of Automotive Engineering, 1 El-Sarayat St., Abbassia, Cairo 11517, Egypt

Abstract: The productivity of wheel-type loader earth moving equipment is highly affected by the digging procedure or style used by the operator. The digging operation could last for more than two shifts per day (16 h), which is very exhausting for many operators. The current technologies of automatic digging algorithm do not completely solve the problem. The single automatic digging procedure feature does not always ensure a good performance, due to the large variety of materials encountered in different sites and even within the same site. In this paper, a fuzzy logic algorithm that detects the different states of the digging operation is proposed. The proposed system is essential for the automatic digging algorithm to capture the operator digging style. During the learning mode, the algorithm determines the key parameters of the system states, which will be then used in the automatic digging algorithm to mimic the same operator procedure. The proposed fuzzy approach was also compared to the conventional Finite State Machine (FSM) approach. The comparison showed that the proposed approach was superior from the state detection point of view.

Keywords: earth moving equipment; FSM; finite state machines; cloned control; behavioural cloning; machine learning; human control behaviour; human control strategy; HMM; hidden Markov model; event-based detection; digging operations; wheel loaders; fuzzy logic; automatic digging algorithm; operator digging styles.

DOI: 10.1504/IJHVS.2013.053010

International Journal of Heavy Vehicle Systems, 2013 Vol.20 No.2, pp.157 - 173

Published online: 30 Oct 2013 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article