鲁棒频域样条优先自适应滤波算法及性能分析

样条自适应滤波结构由线性滤波器和样条插值机制级联组成，是解决Wiener-Hammerstein模型系统辨识的一类有效方案。在非线性系统辨识问题中，随着滤波器阶数增加，将增大时域样条自适应滤波算法的计算复杂度，造成计算效率的降低，且系统附加的非Gaussian噪声会对最小均方算法的样条自适应滤波器性能造成不良影响，导致算法的性能恶化甚至失效。为处理非Gaussian噪声干扰和提高长脉冲响应系统辨识的计算效率，本文结合最大熵准则和频域策略应用于样条自适应滤波器中，并在样条自适应滤波结构中分别采用不同的误差信号对线性部分和非线性部分进行优化，提出了一种鲁棒频域样条优先自适应滤波算法。该算法在滤波前利用非线性系统辨识的不变性原理对未知系统进行优先的有限脉冲响应辨识，可提高非线性系统辨识的精度；通过最大熵准则使算法在非Gaussian噪声环境下具有稳健性，以降低更新过程对大异常值的敏感性；并将线性卷积和线性相关运算通过重叠存储的快速Fourier变换方式进行计算，显著提升了算法的计算效率。此外，本文对所提出的自适应算法进行了收敛性和稳态性能分析，并推导出该算法的理论稳态额外均方误差。最后，通过...     

基于改进的FXLMS算法的主动降噪耳机系统研究

为了有效解决主动降噪耳机系统的低频噪声,采用变步长FXLMS自适应滤波算法,克服传统定步长收敛速度与稳态误差相互制约的不足,通过构造合适的变步长因子,不但改善算法的收敛速度,而且减少了稳态误差.构建主动降噪耳机模拟控制系统,模拟次级通道路径,对噪声进行处理,仿真实验表明,该算法可以有效降低频噪声,相较于传统算法,该算法具有较好的性能.     

鲁棒的高斯和容积卡尔曼滤波红外目标跟踪算法

为提高恶劣测量环境下单站红外搜索与跟踪系统的跟踪性能,提出了一种鲁棒的高斯和容积卡尔曼滤波算法.首先,为改善滤波初值模糊问题,在容积卡尔曼滤波框架下将滤波器分为若干不同初值的子滤波器,利用似然函数逐步减小初值偏差较大的子滤波器权值;其次构建非线性程度判别量,在高非线性情况下将预测密度沿最大特征向量方向进行分割,提高滤波精度;最后利用等价权函数改善新息协方差,减小异常误差对滤波准确性和稳定性造成的影响.实验结果表明,不存在异常误差时,所提算法跟踪结果优于传统算法;存在异常误差时,传统滤波方法的精度明显降低,而所提算法依然能够得到准确可靠的跟踪结果.     

基于Sage-Husa的优化粒子滤波算法

针对非线性系统噪声未知时粒子滤波容易发散或者精度下降的问题,提出一种粒子滤波和改进的Sage-Husa估计器相结合的混合滤波算法。首先用粒子滤波对系统状态进行初步估计,将初步估计值作为次级Sage-Husa滤波器的输入量测值,并与系统状态方程组成新的系统,进而用改进的Sage-Husa算法实时估计系统噪声的统计特性并进行滤波,得到最终的系统状态估计值;为了进一步比较算法的性能,对算法的复杂度进行了定量计算,分析表明优化的算法并未明显提高算法的计算量;最后通过目标跟踪仿真实验验证了算法的有效性。     

多干扰成分的非稳态信号自适应滤波技术

信号在受到色噪声和白噪声混合干扰下产生非稳态突变,导致信号的检测识别性能下降,需要进行滤波降噪处理,提出一种基于格型陷波器级联的非稳态信号自适应滤波技术.构建直接性陷波器,采用高阶检波技术进行滤波器级联设计,构建二阶格型陷波器,根据滤波器传输函数的幅度和相位响应实现线谱增强和噪声抵消,实现多干扰成分的非稳态信号自适应滤波改进.仿真结果表明,采用该滤波算法构建信号滤波器,能有效去除信号中的干扰噪声成分,提高输出信噪比.     

惯性/地磁组合导航系统自适应卡尔曼滤波算法研究

针对惯性/地磁组合导航中遇到的滤波的发散问题,采用自适应卡尔曼滤波估计导航系统的误差.该算法通过实时估计和修正系统噪声以及观测噪声的统计特性达到降低模型误差、抑制滤波发散的目的.在Matlab环境下的仿真证实了该方案可以防止滤波器发散,缩小滤波误差,提高滤波精度.     

单站目标被动跟踪的滤波器方法研究

为了实现对单站目标的被动跟踪,分析并比较了扩展Kalman滤波器和粒子滤波器在非线性估计方面的性能,并且针对粒子滤波器存在的粒子退化现象,引入改进的重采样算法和基于无迹变换的滤波算法.仿真实验分别比较了几种滤波器在目标做匀速、匀加速、变加速情况下距离和速度滤波的均方根误差,结果表明粒子滤波器滤波性能优于扩展的Kalman滤波器,改进的重采样算法和基于无迹变换的粒子滤波器可以有效改善估计精度.     

改善宽带有源消声系统性能的间歇自适应IRLS算法

自适应有源消声控制器常用的算法是滤波-XLMS(FLMS)算法。该算法收敛速度慢,对噪声自相关矩阵特征值散布敏感,应用于宽带有源消声并不十分理想.本文基于最小二乘原理,提出间歇自适应IRLS算法作宽带噪声有源抵消.该算法利用声波从次级源传播到误差传声器的间隔内递推更新自适应滤波器权系数.IRLS算法的特点是收敛速度快,对噪声特征值散布不敏感.仿真结果表明:与采用FLMS算法的消声系统相比,采用IRLS算法的系统在收敛性,稳定性及降噪量等方面均有显著改进。文章最后分析了IRLS算法硬件实现的可行性。     

改进的非数据辅助前向反馈符号定时恢复算法

针对非数据辅助前向反馈(NDA-FF)算法在低滚降系数的成形与匹配滤波器的条件下定时误差大,并且可能出现周期跳转的现象,提出了一种改进的NDA-FF符号定时恢复算法.在对平方定时恢复器的结果进行Kalman滤波的整体结构上,利用两次锯齿波调整,使得估计的延时结果可以直接用于插值.仿真结果表明该算法基本消除了可能出现的周...     

加权融合鲁棒增量Kalman滤波器

在一定环境条件下,当系统的量测方程没有进行验证或校准时,使用该量测方程往往会产生未知的系统误差,从而导致较大的滤波误差。同样地,当系统的噪声方差不确定时,滤波的性能也将会变坏,甚至会引起滤波器发散。增量方程的引入可以有效消除系统的未知量测误差,从而带未知量测误差的欠观测系统的状态估计问题可以转换为增量系统的状态估计问题。该文考虑带未知量测误差和未知噪声方差的线性离散系统,首先提出一种基于增量方程的鲁棒增量Kalman滤波器。进而,基于线性最小方差最优融合准则,提出一种加权融合鲁棒增量Kalman滤波算法。仿真实例证明了所提算法的有效性和可行性。     

Properties of FXLMS-Based Narrowband Active Noise Control With Online Secondary-Path Modeling

Rotating machines such as diesel engines, cutting machines, fans, motors, etc., generate sinusoidal noise signals that may be effectively reduced by narrowband active noise control (ANC) systems. In this paper, a typical filtered-X LMS (FXLMS) based narrowband ANC system equipped with an online secondary-path modeling subsystem is analyzed in detail. First, difference equations governing the dynamics of the FXLMS algorithm for secondary source synthesis and the LMS algorithm for secondary-path estimation are derived in terms of convergence in both mean and mean square. Steady-state expressions for mean-square error (MSE) as well as the residual noise power are then developed in closed form. Extensive simulations are performed to demonstrate the validity of the analytical results.     

基于sigmoid函数的Volterra自适应有源噪声对消器

该文介绍了一种新颖的非线性自适应有源噪声对消器基于sigmoid函数的Volterra自适应有源噪声对消器,并采用输入信号和瞬时误差归一化的LMS自适应算法调整其系数。这种基于sigmoid函数的Volterra自适应有源噪声对消器具有参数少和便于实现的模快化结构等优点。仿真结果表明:这种基于sigmoid函数的Volterra自适应有源噪声对消系统具有良好的抗噪声性能。     

ELMS算法在噪声抵消中的应用

介绍一种改进的算法在噪声消除中的应用.算法利用第二级滤波器估计期望信号,然后在第一级滤波器中消除期望信号对算法误差的影响.仿真表明算法可以明显地降低算法的均方误差,得到很好的输出性能.     

A complex adaptive notch filter using modified gradient algorithm

A modified gradient algorithm is developed for improving the convergence speed of a first-order complex adaptive IIR notch filter, which is used for estimating an unknown frequency of a complex sinusoidal signal embedded in white Gaussian noise. The new cost function using new error criterion is presented and analyzed theoretically. The proposed technique can significantly improve the convergence speed as compared with a complex notch filter using plain gradient algorithm. The computer simulations are conducted to demonstrate the validity of the proposed complex adaptive notch filter.     

时变参数比例自适应滤波算法

在免提电话和视频会议系统中,自适应滤波器估计的回声路径通常是稀疏的.改进的比例归一化最小均方(IPNLMS)算法能够加快自适应滤波器在估计稀疏系统时的收敛速度,但与归一化最小均方(NLMS)算法相比,其稳态失调的波动性较大.为了解决这一问题,本文提出了一种时变参数IPNLMS(TV-IPNLMS)算法.该算法根据系统的均方误差(MSE)与噪声功率的比值,使用一个sigmoid函数来调整时变参数的值.该时变参数能够降低IPNLMS算法在滤波器到达稳态时的比例增益.仿真结果表明,时变参数方法能够降低IPNLMS算法稳态失调的波动性.该算法可用于回声消除、主动噪声控制等领域.     

A new adaptive recursive RLS-based fast-array IIR filter for active noise and vibration control systems

Infinite impulse response filters have not been used extensively in active noise and vibration control applications. The problems are mainly due to the multimodal error surface and instability of adaptive IIR filters used in such applications. Considering these, in this paper a new adaptive recursive RLS-based fast-array IIR filter for active noise and vibration control applications is proposed. At first an RLS-based adaptive IIR filter with computational complexity of order O(n²) is derived, and a sufficient condition for its stability is proposed by applying passivity theorem on the equivalent feedback representation of this adaptive algorithm. In the second step, to reduce the computational complexity of the algorithm to the order of O(n) as well as to improve its numerical stability, a fast array implementation of this adaptive IIR filter is derived. This is accomplished by extending the existing results of fast-array implementation of adaptive FIR filters to adaptive IIR filters. Comparison of the performance of the fast-array algorithm with that of Erikson’s FuLMS and SHARF algorithms confirms that the proposed algorithm has faster convergence rate and ability to reach a lower minimum mean square error which is of great importance in active noise and vibration control applications.     

一种窄带主动噪声控制系统次声学路径建模技术

本文针对最大长度序列相关(Maximal Length sequence Correlation，MLc)建模技术在窄带主动噪声控制系统次声学路径建模的应用中，系统性能易受窄带信号影响这一不足，提出了一种改进的MLC(MMLC)次声学路径建模技术。具体地说就是采用一个自适应预测滤波器来预测和消除MLC技术中的窄带干扰，并用一个补偿滤波器来修正由预测误差滤波器引起的训练信号成分失真。计算机仿真表明，MMLC算法能有效克服窄带主动噪声控制系统次声学路径建模的窄带信号影响，具有较高的建模精度。     

Hybrid filtered error LMS algorithm: another alternative to filtered-x LMS

The filtered-error LMS (FELMS) algorithms are widely used in multi-input and multi-output control (MIMO) active noise control (ANC) systems as an alternative to the filtered-x LMS (FXLMS) algorithms to reduce the computational complexity and memory requirements. However, the available FELMS algorithms introduce significant delays in updating the adaptive filter coefficients that slow the convergence rate. In this paper, we introduce a novel algorithm called the hybrid filtered-error LMS algorithm (HFELMS) which, while still a form of the FELMS algorithm, allows users to have some freedom to construct the error filter that guarantees its convergence with a sufficiently small step size. Without increasing the computational complexity, the proposed algorithm can improve the control system performance in one of several ways: 1) increasing the convergence rate without extra computation cost; 2) reducing the remaining noise mean square error (MSE); or 3) shaping the excess noise power. Simulation results show the effectiveness of the proposed method.     

Feedback active noise control based on transform-domain forward–backward LMS predictor

In this paper, a new feedback active noise control (FBANC) system based on the transform-domain forward–backward LMS (TFBLMS) predictor has been proposed. The new ANC system employs the TFBLMS predictor for its main-path (MP) predictor as well as for the noise canceller. To overcome the ill effect of the primary noise field, which acts as an observation noise for the secondary-path (SP) identification, the noise canceller is used. As the main-path predictor is based on the TFBLMS, its convergence rate improves due to its input orthogonalization. Further, its FBLMS nature reduces misadjustment. The use of TFBLMS predictor for noise canceller also gives a good prediction of primary noise at a faster rate, enabling improved SP identification. This improved SP identification indirectly aids the MP predictor to achieve an improved performance. A new filtered-x LMS structure has been proposed to realize the new MP predictor to accommodate the TFBLMS algorithm. The TFBLMS algorithm is applied directly to the noise canceller for SP identification. The proposed new ANC system has been found to have a significantly better noise reduction (by 14.6 dB) over the FBANC system based on tapped delay line time-domain FBLMS algorithm.     

Robust M-estimate adaptive filtering

An M-estimate adaptive filter for robust adaptive filtering in impulse noise is proposed. Instead of using the conventional least-square cost function, a new cost function based on an M-estimator is used to suppress the effect of impulse noise on the filter weights. The resulting optimal weight vector is governed by an M-estimate normal equation. A recursive least M-estimate (RLM) adaptive algorithm and a robust threshold estimation method are derived for solving this equation. The mean convergence performance of the proposed algorithm is also analysed using the modified Huber (1981) function (a simple but good approximation to the Hampel's three-parts-redescending M-estimate function) and the contaminated Gaussian noise model. Simulation results show that the proposed RLM algorithm has better performance than other recursive least squares (RLS) like algorithms under either a contaminated Gaussian or alpha-stable noise environment. The initial convergence, steady-state error, robustness to system change and computational complexity are also found to be comparable to the conventional RLS algorithm under Gaussian noise alone