平滑l_1模递推辨识算法

Recursive Identification Algorithm with Smoothed l1 Norm Criterion

通过对ｌ１模指标的平滑处理，构造了平滑近似指标下的递推辨识算法．这种算法实时计算负担小，且具有良好算法性质．数值仿真结果表明，它仍蕴含原指标意义下的鲁棒估计性质．

A new smoothed absolute error criterion (smoothed l1 norm) for constituting robust recursive identification algorithm is proposed. The presented smoothed criterion leads to less complexity for numerical computing than original l1 norm does. The properties of presented algorithm are discussed. To show the effectiveness of the algorithm, two examples are given. Via simulation, it is demonstrated that the algorithm yields a good performance even in the presence of strong disturbances.