加入前馈补偿的时滞系统鲁棒性跟踪控制器

Robust Tracking Controller of Time Delay System with Feedforward Compensation

时滞系统的优化跟踪控制在过程控制、飞行控制等领域应用广泛,传统的Smith控制对时变二自由度时滞系统控制效果不好,偏差较大时的控制鲁棒性不高。为了提高系统的稳定性和快速反应能力,在对时滞系统鲁棒性跟踪控制器进行改进设计的基础上,提出了一种加入前馈补偿的时滞系统鲁棒性跟踪控制器设计方法。应用该方法基于内模控制思想构建了二自由度的时滞控制系统,采用常规Smith控制器结构实现被控对象的输入输出描述,在经典的Smith控制器设计的基础上,加入前馈补偿算子自适应训练控制器神经元,采用最速下降法构建LMBP算法,避免了时滞系统的鲁棒性跟踪控制的最优值陷入局部收敛,实现了对时滞系统鲁棒性跟踪LM-Smith控制器的改进设计。仿真结果表明,采用该方法的时滞系统控制精度较高,鲁棒性较好,控制性能有明显改善。

Optimal tracking control for time delay systems are widely used in process control and flight control. The effect of traditional Smith control in time-delay system with time-varying two degrees freedom is not so good that the control robustness is not satisfied with big deviation. In order to improve the stability and fast response ability of the system, design method for robust tracking control of time delay systems has been presented on the basis of improvements on robust tracking controllers of time-delay systems with feedforward compensation and thus two-freedom-degree time-delay systems has been established with internal model control structure to describe the input and output of controlled objects with conventional Smith controller＇s structures. Based on the design of Smith controller,the feed forward compensation controller training operator adaptive neuron has been introduced into the time-delay systems and the LMBP algo- rithm has been constructed with steepest descent method. Therefore the local convergence of optimal robust tracking control value in time -delay systems has been avoided with the implementation of improvements on robust LM-Smith controller design of time-delay system. The simulation results show that the robustness of this proposed method is better than other methods and its control accuracy is higher with less error and that the performances of the established time-delay systems have been promoted significantly.