基于PI^λD^μ控制器的一种线性系统辨识方法研究

Identification Algorithm for Linear System Based on Fractional-order PI^λD^μ Controller

近年来,分数阶控制研究愈来愈被关注。传统系统参数辨识算法主要基于在某种意义下的最优数据拟合思想,往往比较复杂、计算量大。我们将辨识误差的抑制问题,看作是以误差方程来描述的系统的状态渐近稳定性来处理,从而将反馈控制的思想引入辨识算法中,提出了一种基于分数阶PIλDμ控制器的线性系统参数时域辨识算法,给出了递推算法的详细推导,并用已知系统仿真验证了算法的有效性。该算法简便、计算量小。仿真结果表明,该算法优于传统最小二乘法。

The fractional-order control has drawn much research attention in the field of science and technology of control recently. The conventional identification is founded in the data fitting and is usually complicated and high-computational. The correction of the identification error was regarded as the state stabilization problem of the system described by the error equation. A new method for time domaim identification of linear system based on Fractional-order PI^λD^μ controller was proposed. The recursive algorithm was deduced in detail, and its effectiveness was verified by two simulation examples of known systems. The operation is easy and the amount of calculation is decreased. The results of simulation indicate that this method is better than the traditional least squares algorithm.