互联网环境下基于预测控制的闭环控制方法

Predictive control-based closed-loop control under the internet environment

针对基于互联网的控制系统中,网络数据传输普遍存在随机时延、数据时序颠倒和数据丢失而难以实现闭环实时控制的问题,提出一种基于预测控制思想的闭环实时控制方法.该方法通过对采样数据附加时间标志,选择和利用当前时刻所能得到的最新输出测量值进行在线校正,并在互联网上实现预测控制策略.在模型准确或模型失配情况下,控制效果与无网络传输环节时接近.仿真结果表明了该方法的有效性.

Closed-loop real-time control over the internet is very difficult because of the commonly existed random time delay, data packet dropout, and data packet out of order during internet transmission. A new strategy based on predictive control is proposed to deal with these problems. By attaching timestamp to the sampled data, choosing the available newest output measurement, and using this output measurement to modify the predicted outputs in predictive control, the strategy is able to effectively overcome the problems mentioned above. In both case of without and with model mismatch, the control performance is comparable to that of with no network transmission. The simulation results show the effectiveness of the scheme.