网络控制系统的变采样周期调度算法

综合考虑网络控制系统的误差、误差变化率、网络利用率及采样周期对系统性能的影响，设计了一种基于模糊反馈的变采样周期调度算法。该算法由网络利用率预测和采样周期调节两部分组成：网络利用率预测部分根据当前网络运行状况预测新的网络利用率；采样周期调节部分包含网络利用率分配和采样周期的计算。采样周期调节部分的网络利用率分配，用于重新分配各控制回路的网络利用率，分配时考虑系统各回路的误差和误差变化率，利用模糊控制理论调整各回路对网络的需求程度，完成分配；而采样周期计算是根据所得的网络利用率及数据的传输时间，动态调节系统各回路的采样周期。最后，结合EDF调度算法利用TrueTime工具箱对所研究的调度算法进行了仿真，结果表明采用本文所研究的变采样周期调度算法的控制系统性能要优于采用固定采样周期调度算法的控制系统性能。

In this paper, considering the effects on system performance occurred by system error, error rate, network utilization and sampling period, we design a new variable sampling period scheduling algorithm based on fuzzy feedback. The algorithm consists of network utilization prediction and sampling period adjustment. In the network utilization prediction procedure, a new network utilization is predicted according to real-time network operation status; Network utilization allocation and sampling period calculation are done in the sampling period adjustment procedure. Network utilization allocation is used to allocate the network utilization of each loop considering system error and error rate, and the allocation is conducted through using the fuzzy control theory to adjust the demand degree of each loop. According to the network utilization and data transmission time, the sampling periods of loops are adjusted dynamically through the sampling period calculation. Finally, the proposed algorithm is simulated using True Time toolbox to compare with the fixed sampling period scheduling algorithm. The simulation results show that the former has better control performance than the latter.