Real time scheduling for autonomous mobile robots using genetic algorithm and constraint based reasoning

Abstract

In many real-time systems, relative timing constraints are imposed on a set of tasks. Generating a correct ordering for the tasks and deriving their proper start-time assignments is a Nondeterministic Polynomial Problem (NP) hard problem; it subsumes the Non-preemptive Scheduling Problem. Real-time systems are characterized by the presence of timing constraints on the computations carried out by the system. Examples are found in domains such as avionics, process control and robotics, where a computer is used to control and manipulate a physical system. Real-time systems are characterized by computational activities with timing constraints and classified into two categories, hard real-time system and soft realtime system. In hard real-time system, deadline missing can be catastrophic. However, in the case of soft real-time systems, slight violation of deadlines is not so critical. Autonomous mobile robot system is one of typical task scheduling of soft real-time system. In this study we addressed the problem of real-time scheduling in autonomous robot and in particular we compared two different scheduling approaches, hybrid GA and Pre-emptive Rate Monotonic. The results have shown that in this case a hybrid GA approach is preferable because it is more efficient and moreover the disadvantages of such a choice have shown not to be relevant to the overall functioning of a typical autonomous robot application. In fact, the greater efficiency could be exploited to minimize the overall deadline missing among the tasks and therefore robot can work smoothly.