Real-time Control System, Multi-task Fuzzy Dynamic Scheduling And Parameter Scheduling Method

RCS(Real-time Control System) has been paid more and more attention in recent years, and became a significant part of information industry, such as aeroplane control, rocket control, submarine control. And the scheduling method, which distributes tasks to the computer, exerts a significant influence on performance of RCS.EDF (Earliest Deadline First) as the best scheduling algorithm has been drawn great attention. However, its control performance will become poor when the system is overload. Therefore a new scheme based on EDF algorithm is proposed, which designed appropriate scheduling algorithm and controller according to tasks' dynamic execution condition. And a multi-loop RCS system model with schedules is built based on TrueTime. In order to improve the performance of control system with EDF, a new scheduling algorithm-FEVF(Earliest Fuzzy Value time First) was proposed, which combined fuzzy theory with EDF and then applied to RCS model based on TrueTime. Based on control performance cost and guarantee ratio, computer simulation are given to demonstrate the effectiveness of the proposed FEVF algorithm, thereby showing whose superiority to the conventional EDF and more suitable for control system.Furthermore, considering the drawback of classic HPV(hybrid priority and value) algorithms, this paper proposed a new scheduling scheme based on priority table design by integrating two characteristic parameters(i.e. value degree and Importance) of a task. Two real-time scheduling algorithms from the scheme are presented: VPT(value priority table) and PVT (priority value table). Based on control performance cost and guarantee ratio, the performance of the HPV, VPT and PVT algorithms are analyzed. The experimental results shows that the VPT algorithms owns a better performance compared to HPV(hybrid priority and value) and VPT algorithms under overload conditions.