RESOURCE SCHEDULING STRATEGIES FOR A NETWORK-BASED AUTONOMOUS MOBILE ROBOT

Abstract: TIn this paper, efficient resource scheduling strategies for a network-based autonomous mobile robot with hybrid robot control architecture are proposed. The basis of the resource scheduling strategies is the DMS(Deadline Monotonic Scheduling), and soft deadline and hard deadline are newly defined wit the DMS. A coordinated scheduling strategy configures the soft deadlines considering proceeding constraints for the minimization of behavior-to-behavior delays. A modulation strategy of the hard deadlines-rates is also proposed in this paper for the optimal usage of computational resource. The modulation is performed based on the dynamical values of the behaviors running on an operating system in each node. To compensate the uncertainty of the dynamical values inferred by sensor information, the fuzzy controls are used for the modulation, and the genetic algorithm is also used to refine the rate modulation result considering optimal real-time feasibility. Through simulation tests, scheduling results with the proposed strategies are compared with results without the proposed strategies to verify the performance enhancement by the proposed strategies. TCopyright © 2005 IFA