基于加权递推最小二乘法的微硬盘伺服信号滤波

针对微硬盘伺服信号的滤波问题,采用了一种加权递推最小二乘法.该算法使用最小二乘加权时间平均准则取代通常的最小均方误差准则,并利用该准则导出权向量的自适应迭代公式,算法保证在每步迭代运算中权值都能调整到最佳,同时降低了权值调整量对噪声的敏感程度.实验结果表明:该算法滤波效果好、抗噪能力强、稳态收敛精度高和收敛速度快,适用于诸如微硬盘驱动器之类对收敛速度及消噪要求较高的场合.

Filtering of the servo signals from micro-HDD by weighted recursive least square algorithm

In order to solve the problems of signals filtering in micro hard disk drive(Micro-HDD),used was weighted recursive least square(WRLS) algorithm,in which least mean square(LMS) error principle was substituted by the rule of minimizing weighted mean of total square error from initial time to current time and the adaptive iterative formula of weight vector was derived to ensure optimal weight value in every iterative operation and reduce the sensitivity of the weight adjusting value to noise.Experimental results show that this algorithm displayed its excellent filtering effect,anti-noise performance,stable convergence precision and speed,being suitable for Micro-HDD in the faster convergence speed and higher noise attenuation.