基于负载均衡的MVB周期数据调度优化  

Optimization of MVB Periodic Data Scheduling Based on Load Balancing

作　　者：王保华[1] 严翔[2] 王立德[1] 王昕[1] 

机构地区：[1]北京交通大学电气工程学院,北京100044 [2]中国铁道科学研究院机车车辆研究所,北京100081

出　　处：《中国铁道科学》2017年第5期114-120,共7页China Railway Science

基　　金：国家自然科学基金资助项目(61401017);"十二五"国家科技支撑计划项目(2015BAG13B01-02)

摘　　要：针对采用单调速率(RM)调度算法构建的周期扫描表易导致多功能车辆总线(MVB)网络出现负载不均衡的问题,基于MVB网络建立以网络负载不均衡度最小为目标的优化模型;为增加算法的收敛速度,简化算法求解过程,采用基于以平均负载率为中心的预测域的免疫遗传算法对优化模型进行全局最优求解。仿真和试验结果表明:较RM调度算法和其他优化算法,采用所给出的优化模型和算法能够明显减小MVB网络负载的波动;算法具有更快的收敛速度,可以提高MVB网络的负载能力,提升控制系统的性能。Aiming at the problem of load imbalance in multifunction vehicle bus（MVB）network caused by the timed period polling table made by the rate monotonic（RM）scheduling algorithm,an optimization model with the target of minimum network load imbalance was established based on MVB network.In order to increase the convergence speed of the algorithm and simplify the algorithm,the immune genetic algorithm based on the prediction domain with the average load rate was used to solve the optimal model globally.Simulation and experiment results show that the optimal model and algorithm can significantly make the fluctuation of MVB network load smaller than RM scheduling algorithm and other optimization algorithms.The algorithm has faster convergence speed and can improve the load capacity of MVB network and improve the performance of control system.

关 键 词：列车控制网络 多功能车辆总线 负载均衡 免疫遗传算法 预测域 负载率 

分 类 号：U284.482[交通运输工程—交通信息工程及控制]