Optimal Scheduling Algorithm for Multi-Tasks in Distributed Control Systems

There are real-time periodic tasks and non-real-time sporadic tasks in Distributed Control System (DCS) and how to schedule these tasks influences not only on the resource utilization of system, but also on the control performance of system. Firstly, task models and system model are given in this paper. A scheduling algorithm called dual-priority queue scheduling algorithm is presented for uniprocessor and the schedulable condition is given. Based on this, the task scheduling algorithm for DCS is investigated based on heuristic task allocation method. According to the relationship of the performance of DCS and sampling periods of periodic tasks, particle swarm optimization (PSO) algorithm is applied to optimize the performance index of DCS. Simulation results show that the performance index of DCS can be improved obviously in the case that the schedulability of all tasks is guaranteed by adopting the algorithm presented in this paper.