基于时延预测模型的网络补偿控制器设计*

针对网络控制系统中随机双向通道时延引起的不确定性问题,提出一种基于时延在线预测模型改进的补偿控制方法。首先,基于自回归模型构建时延预测模型,通过递推最小二乘法实时更新模型参数以预测前向通道时延,并由区间分割法获得高精度的环回时延；然后,设计状态预估器、网络预测器和时延补偿器对环回时延进行补偿,进一步提高系统性能。采用TrueTime工具箱构建网络闭环仿真系统以仿真实际的网络传输情况,同时对PI控制、传统的时延补偿控制以及改进后的时延补偿控制进行对比实验,验证该方法的优越性。

The design of networked compensation controller based on the delay prediction model

Aim at the issue of the uncertainty caused by the time delay in the two channels of networked control systems, an improved compensation control method based on the online delay prediction model is proposed. First of all, the delay prediction model is built based on regression model and the model parameters are updated in real time by the recursive least squares method to predict the forward channel delay. And the interval partition method is used to obtain high precision of the loop delay. Then, the state predictor, the network predictor and the delay compensation is designed for loop delay compensation to further improve the system performance. Using TrueTime toolbox to construct the simulation system, the actual network transmission is simulated and the comparison of PI control, the time delay compensation control and the improved compensation control is taken to verify the superiority of the method.