基于AR模型时延预测的改进GPC网络控制算法

Improved GPC network-control algorithm based on AR model time-delay prediction

针对网络随机时延等不确定因数导致网络控制系统的控制性能下降甚至不稳定,提出一种新的网络控制算法.首先建立网络时延的自动回归模型;然后采用参数自校正的最小均方算法对网络时延进行在线预测;最后采用一种考虑时延的改进广义预测控制算法对网络时延进行补偿.仿真实验结果表明,该方法对网络时延具有较好的补偿效果,且在线计算量小,有很好的实时性,同时对干扰和丢包等因素具有良好的鲁棒性.

In the network control system, uncertain factors such as network delays can usually deteriorate the control performance and stability of networked control systems. Therefore, a new network control algorithm is proposed. First, the auto regressive model for network time delays is established, and parameter self-tuning LMS algorithm is used for online prediction. Then improved generalized predicive control based on state-space model is adopted to compensate the time- delay. Finally, simulation results show that this method has a good effect on time delays compensation and has a good real-time because of the small amount of on-line calculation, at the same time, the method has good robustness according to interference and data loss.