A Proposed Methodology for Human Operator Modeling and Development of a Virtual Operator Model for Control of a Simulated Compact Track Loader

Résumé fourni par la source

A methodology for performing Human Operator Modeling (HOM) using a Caterpillar Model 299D3 XE Compact Track Loader (CTL) is presented. The proposed method uses task analysis techniques to decompose material excavation and moving tasks into smaller, individual tasks presented in a task list. A method for verifying and refining the task list is presented, along with a procedure for identifying relevant human operator sensory information and analyzing human decision making in the context of CTL operation. This methodology is then partially verified through the analysis of a non-expert human operator in Vortex Studio, a realistic construction equipment simulator. A modified test course is executed by a non-expert human operator in the simulation environment, and the recorded data is used to create a quantitative Human Operator Model. From this, a Virtual Operator Model (VOM) feedback controller simulating the performance of the human operator is developed. The VOM is implemented using a state machine to transition between individual tasks. Fuzzy Logic Control (FLC), is implemented for each task to control bucket tilt, arm lift, and throttle, with controller parameters calculated from the quantitative HOM data. The VOM controller is verified using the same test course performed by the human operator. The performance of the human operator is compared to that of the VOM controller in order to validate the HOM and VOM methodology for a simulation environment.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.