ملخص
This work develops distributed hierarchical task assignments and communication algorithms for flocks of unmanned aerial vehicles dispersed in an unknown theater while engaging multiple moving targets. The dynamical changes require distributed algorithms, without relying on a central agent, which may constitute a single point-of-failure and is exposed to communication breaks. Our methodology overcomes the typical deficiencies of a centralized solution by organizing the agents in spanning trees. Whenever possible, the spanning trees merge into a single tree that clusters the maximum number of agents. Using relaxation methods and inputs from its parent and children, every agent attempts to optimize its own solution for task assignment while communicating using an ad-hoc protocol. Simulation experiments show clear-cut advantages of using the proposed task assignment and communication algorithms.
اللغة الأصلية | الإنجليزيّة |
---|---|
الصفحات (من إلى) | 234-250 |
عدد الصفحات | 17 |
دورية | Journal of Aerospace Computing, Information and Communication |
مستوى الصوت | 5 |
رقم الإصدار | 8 |
المعرِّفات الرقمية للأشياء | |
حالة النشر | نُشِر - أغسطس 2008 |
منشور خارجيًا | نعم |