基于延时预测的模型状态反馈网络控制方法

Model States Feedback Networked Control Method Based on Delay Prediction

网络延时的存在不仅会使控制系统的动、静态性能变差,还可能会破坏系统的稳定性,造成既有控制方法的失效。针对带有随机延时的网络控制系统,提出了一种基于网络延时实时预测和系统状态逼近的控制方法。首先,建立无延时的理想控制系统模型作为标准控制系统模型;其次,应用基于隐马尔可夫模型(HMMs)的延时预测算法预测网络延时,该方法能够实时自适应学习,以适应网络负载情况的变化,得到更符合实际情况的延时预测值;最后,根据网络延时的预测值、实际网络控制系统与标准控制系统模型的状态之差,来调整控制信号,从而使实际网络控制系统的状态与标准控制系统模型接近、甚至相等,达到期望的控制效果。数字仿真结果表明,这一方法是可行的,能够获得更加快速的动态性能和稳定的静态性能。

The network-induced time delay can not only degrade system performances but also cause system instability, which may in- validate the existing control methods. Aiming at networked control systems with random time delay, a novel networked control method is proposed based on the real-time prediction of network-induced time delay and the system states approaching. First of all, an ideal model of the control system without network-induced time delay is built as a standard model of control system; Secondly, a network-induced time delay model using hidden Markov models （HMMs） is introduced to predict network-induced time delay, which adopts adaptive learning to adapt to new situations of network load in order to get a more realistic prediction of the network-induced time delay; Finally, the control signal in the current sampling period is adjusted according to the predicted network-induced time delay value and the differ- ence in the states between ideal model of control system and actual networked control system in order that the states of actual networked control system are closed to the states of ideal model of control system, the desired control performances are achieved. The digital stim- ulation results prove that the method presented is feasible and the dynamic and static resoonse oerformances are satisfied.