An improved spline adaptive filter for nonlinear system identification under impulsive noise environment

Abstract

Aiming at the negative influence of impulse noise in conventional spline adaptive filtering (SAF) algorithm in nonlinear system identification, an improved spline adaptive filtering algorithm is proposed. The algorithm mainly improves on two aspects. First, the cost function is constructed by using the hyperbolic tangent function. When the independent variable fluctuates sharply, the cost function will only approach a finite value, which improves the robustness of the filter in an impulse noise environment. In addition, a pre-filtering observation strategy is proposed to further improve the convergence speed of the algorithm and reduce the steady-state error caused by the iterative process in the system. Through the analysis of the amplitude of the noise in the time domain and the frequency domain, the update rule of the weight is determined. The simulation results show that it has better performance than the existing SAF algorithm.