分布式系统TTCAN调度优化算法研究

Research on TTCAN scheduling optimization algorithm for distributed systems

摘要:

总线技术的发展给线缆测试仪带来了分布式、信息化、网络化的新需求，且在分布式线缆测试仪工作过程中，测试线路的数目增加也对总线数据通讯的稳定性和通讯效率提出了更高的要求。针对分布式系统在线缆测试中的应用需要，设计并优化了分布式线缆测试仪工作的TTCAN应用层协议和其系统矩阵。对于分布式系统通信中的周期性消息形成的系统矩阵先后采用遗传算法、改进型差分进化算法进行优化，对于其中的非周期性消息采用基于松弛度的动态优先级算法。在MATLAB仿真环境中进行实验，实验结果表明，改进型差分算法比遗传算法能够更快、更稳定地计算出优化矩阵，经调度优化后的TTCAN总线工作时数据传输效率有显著提高。论文通过智能优化算法，有效提高了系统总线的通讯效率和稳定性。

Abstract:

The development of bus technology brings new requirements of distributed, information and network to the cable tester, and the increase of the number of test lines in the working process of the distributed cable tester also puts forward higher requirements on the stability and communication efficiency of the bus data communication. According to the application requirements of distributed system in cable testing, the TTCAN application layer protocol and its system matrix of distributed cable testing instrument are designed and optimized. The system matrix formed by periodic messages in distributed system communication is optimized by genetic algorithm and improved differential evolution algorithm successively, and the dynamic priority algorithm based on relaxation degree is adopted for the non-periodic messages. Experiments in MATLAB simulation environment show that the improved difference algorithm can calculate the optimization matrix more quickly and stably than the genetic algorithm, and the data transmission efficiency of TTCAN bus after scheduling optimization is significantly improved. In this paper, the communication efficiency and stability of the system bus are improved effectively by the intelligent optimization algorithm.