A group of autonomous agents can agree on a predetermined set of social laws, inorder to reduce the complicity of conflict resolution later in the field when they are dispatched to achieve the given task. However, it is conceivable that under certain situations, the agents may have to violate the social jaws so that they will achieve the task more efficiently or even survive in the field. We thus consider the issue of fusing the social jaws and a set of super rules for handling exceptions. We report some preliminary results in the context of dynamic path planning for a group of autonomous mobile robots. It is conceivable that when a group of autonomous robots are dispatched to achieve the given task in the field, they can coordinate their motion and resolve conflicts more efficiently if they have a predetermined set of social laws that they obey (3). However, simply obeying the social laws strictly all the time may not be an ideal strategy for the robots, since doing so might require them to take some actions that turn out to be unnecessary in certain situations. So in this paper, we investigate the possibility of letting the robots plan their motion more adaptively by allowing them to override the social laws occasionally.(KAR) P. 1.
展开▼
