中文摘要：

针对多回路的网络控制系统,本文同时考虑系统误差和误差变化率,设计了一种基于神经网络的模糊动态调度算法.该算法根据系统中各回路的误差和误差变化率,利用神经网络模糊控制的方法实时调整各回路的优先级,从而实现对网络控制系统的调度.最后,利用TrueTime工具箱建立了包含模糊动态调度器的网络控制系统仿真模型,并将其与RM和EDF调度算法进行对比.仿真结果表明,在相同的网络带宽占用条件下,本文所设计的模糊动态调度算法相比于RM和EDF调度算法,产生的网络诱导时延更小,且具有较好的控制性能.

英文摘要：

For a multi-loops Networked Control System （NCS）, considering both system error and error rate, the design of a fuzzy dynamical scheduling algorithm based on Neural network was presented in the paper. According to the system error and error rate of each loop, the algorithm uses the neural network method to adjust the loops＇ priority real-timely, so realize the scheduling of NCS. Finally, TrueTime Toolbox is used to model the NCS which contains a fuzzy dynamical scheduler, and compared it with RM and EDF dispatch algorithms. The simulation results shown that, in comparison to RM and EDF dispatch algorithms, the designed fuzzy dynamical scheduler provides a smaller networkinduced delay and better control performance.