如何调度这些信息不仅影响网络利用率,而且影响系统的控.doc

摘要：网络控制系统是以网络为基础，实现资源共享和协同控制的分布式实时系统。系统中有多种信息需要在网络上传递，如何调度这些信息不仅影响网络利用率，而且影响系统的控制性能。因此，信息调度技术是网络控制系统设计和实现的关键技术之一。本文针对网络控制系统中存在的信息设计调度算法。 首先，针对网络控制系统中的周期信息，本文给出了信息模型和分析了信息之间的相互关系。基于此，采用了非抢占EDF 算法调度进行信息调度，并给出了该算法基于令牌总线的实现方法。分析了信息的可调度性，给出了信息可调度性的判断条件。根据各子信息之间存在的相互关系，对子信息时限进行了优化，以提高网络的利用率。仿真结果表明本文给出的方法是有效的。 最后，对于同步采样的网络控制系统中的周期与非周期信息，分析了信息的特点，给出了信息模型。在此基础上，通过设计调度序列实现信息调度。考虑到控制系统的性能与采样周期有关，采用整数编码为基础遗传算法对调度序列进行优化，以改善系统的控制性能。 关键词：网络控制系统；信息调度；调度算法；EDF 算法；优化时限Abstract: Networked Control System NCS is network-based distributed real-time system to realize resource sharing and collaborative control. There are many kinds of messages which need to be passed on the network. How to schedule these messages not only affects the network utilization, but also affect the control performance of the system. Therefore, the message scheduling technique is one of key technologies to design and implement networked control systems. In this dissertation, scheduling algorithms for messages in NCS are designed. Firstly, for the periodic messages in NCS, message model is given and the relationship between the messages is analyzed. Based on this, non-preemptive EDF algorithm is adopted to schedule messages, and the method to implement this algorithm on the basis of the token bus. The schedulability of the algorithm is analyzed and schedulable condition is investigated. According to the mutual relationship between sub-messages, deadlines of sub-messages have been optimized to improve network utilization. The simulation results show that the method presented in this dissertation is effective. Finally, for the periodic messages and sporadic messages in the synchronous networked control systems, their characteristics are analyzed and the message models are given. Basing on this, scheduling sequence is designed to implement to schedule messages. Considering that the performance of control systems is influenced by the sampling period. The scheduling sequence is optimized by adopting genetic algorithm