Task planning and behavior scheduling for a Reconfigurable Planetary Robot System

In this paper, the synthetical approach of behavior tree and logic syntactic representation is presented to implement task planning and behavior scheduling for autonomous motion, reconfiguration and cooperation in a new planetary rover, Reconfigurable Planetary Robot System (RPRS). This approach provides support for task decomposition, behavior scheduling, synchronization, execution monitoring and exception handling. The analytic process is illustrated in an example. The modularized design method of single child-robot is proposed and implemented. Two-layered hierarchical hybrid control architecture is designed for RPRS, which applies the centralized plan at task-level and the distributed control at motion-level. As a result, child-robots can reconfigure to some configurations and perform some tasks cooperatively. The latest results of this research are demonstrated in the experiments of the prototype.