A fast RLS algorithm for second-order Volterra filter is analyzed. And the contradiction between speed and accuracy of convergence is revealed. The modified RLS algorithm is presented by substituting constant forgetti...
详细信息
A fast RLS algorithm for second-order Volterra filter is analyzed. And the contradiction between speed and accuracy of convergence is revealed. The modified RLS algorithm is presented by substituting constant forgetting-factor with function of forgetting-factor constructed in this paper. The rules for constructing forgetting-factor function and the method for selecting parameters is discussed. Finally, a numerical example shows that the modified RLS algorithm result in faster convergence speed by solving the contradiction between speed and accuracy.
In this paper we apply a filtered-X algorithm to an active feedback control structure and derive the transfer function of a closed-loop control system. Simulation studies are then carried out on the closed-loop proper...
详细信息
In this paper we apply a filtered-X algorithm to an active feedback control structure and derive the transfer function of a closed-loop control system. Simulation studies are then carried out on the closed-loop property while varying the parameters (input frequency, delays in plant, amplitude and phase of modeling filter). Several properties of adaptive feedback control are revealed. Experimental studies on feedback active noise control of noise in a finite duct and a small enclosure are described, and outstanding active noise control effects are achieved. Experimental results of closed-loop frequency response are also provided.
作者:
Lim, JSAjou Univ
Div Informat & Comp Engn Paldal Gu Suwon 442749 South Korea
The linearly filtered gradient least mean square algorithm is reformulated using exponentially weighted least-square errors, and a partially filtered gradient LMS algorithm is proposed as a variant. It is shown that t...
详细信息
The linearly filtered gradient least mean square algorithm is reformulated using exponentially weighted least-square errors, and a partially filtered gradient LMS algorithm is proposed as a variant. It is shown that the proposed algorithm is superior in terms of convergence and tracking capability compared to the LMS and FGLMS algorithms.
作者:
Lim, JSAjou Univ
Grad Sch Informat & Commun Suwon 442749 South Korea
In this letter, we propose tno algorithms for adaptive finite impulse response (FIR) filters as variants of the filtered gradient adaptive (FGA) algorithm. The proposed algorithms are formulated from the FGA by introd...
详细信息
In this letter, we propose tno algorithms for adaptive finite impulse response (FIR) filters as variants of the filtered gradient adaptive (FGA) algorithm. The proposed algorithms are formulated from the FGA by introducing an orthogonal constraint between the direction vectors, The algorithms employ the forgetting factor optimized on a sample-by-sample basis so that the direction vector is orthogonal to the previous direction vector, while the FGA algorithm uses a fixed one. It is shown through the computer simulations that the proposed algorithms are superior in terms of convergence and tracking capability to the FGA algorithm.
暂无评论