工业过程抗非重复性干扰改进的迭代学习控制

Improved Iterative Learning Control for Resisting the Non-repetitive Disturbance in Industrial Process

对于实际工业过程系统中存在的非重复性干扰,传统的PD型迭代学习控制不能很好地加以抑制。为此,提出加权PD型指数变增益加速闭环迭代学习控制算法。通过采集非重复性扰动信号,将其转化为设定值阶跃变化的序列,并采用改进的加权PD型指数变增益闭环算法,消除非重复性干扰,从而获得更为理想的系统输出,使控制系统的动态性能得到改善。算法研究表明,当迭代次数趋于无穷时,跟踪误差一致收敛到零。系统仿真验证了所提控制算法的有效性。

Non-repetitive disturbances often exist in practical industrial process systems, while traditional PD-type iterative learning control cannot restrain these well. Thus the iterative learning control algorithm with weighted PD-type exponential variable gain closed-loop is proposed. Through collecting non-repetitive disturbance signals and converting these into the sequence of step change of set-point, and adopting improved weighted PD-type exponential variable gain closed-loop algorithm, non-repetitive disturbances are eliminated to obtain ideal system output and better dynamic performance of the control system. The verification of convergence proves that the tracking error converges to zero when the iteration goes towards to infinite. The system simulation has proved the effectiveness of the proposed control algorithm.