关于ISO2631 滤波计权的问题，如何从时间域内对其进行滤波计权，数字滤波器如何设计？

这个是可以在matlab中实现，但是相当的麻烦，而且执行效率低。方法可以说明一下：因为ISO2631里已经有相关模拟滤波器的参数，那么你要做的就是如何将这个模拟滤波器转换成数字滤波器，在matlab中可以用双线性变换将其变换成数字滤波器，bilinear()函数。再利用filter进行滤波，本人已经将程序调通，对照频域法滤波效果一样。

为大家提供一个简单的方法，就是利用famos谱分析套件中带有悔洞的滤波函数，实现滤波。只要将慎尺要滤波的变量写入，并填碧孝枯写相关参数即可得到滤波后的结果。

VibrationFilter(InputChannel, FrequencyWeighting, TimeConstant, Reduction)

参数介绍

InputChannel:所以进行频率滤波加权的变量

FrequencyWeighting:频率计权的选择：

10:Wk, z direction and for vertical recumbent direction, except head. As per ISO 2631-1, 2nd edition, 1997

11:Wd, x and y directions and for horizontal recumbent direction. As per ISO 2631-1, 2nd edition, 1997

12:Wf, motion sickness. As per ISO 2631-1, 2nd edition

13:Wc, seat-back measurement. As per ISO 2631-1, 2nd edition, 1997

14:We, measurement of rotational vibration. As per ISO 2631-1, 2nd edition, 1997

15:Wj, vibration under the head of a recumbent person. As per ISO 2631-1, 2nd edition, 1997

16:Hx, whole body vibration, standing, sitting position: x- and y-directions. Recumbent position: y- and z-directions. As per DIN 45671 Part 1, Sept. 1990.

17:Hz, whole body vibration, standing, sitting position: z-direction. As per DIN 45671 Part 1, Sept. 1990.

18:Hxl, whole body vibration, recumbent position: x-direction. As per DIN 45671 Part 1, Sept. 1990.

19:Hb, whole body vibration, body posture not specified. As per DIN 45671 Part 1, Sept. 1990.

20:Hh, hand-arm-vibration, for all directions. As per DIN 45671 Part 1, Sept. 1990.

20:hand transmitted vibration, weighting filter. As per ISO 7505, 1st edition, 1986-05-01.

TimeConstant: The time constant for finding the exponentially weighted RMS. Specified in seconds, >= 0.0. E.g. 1.0s for SLOW weighting, 0.125s for FAST weighting. If = 0, no RMS value calculated. In this case, only the (band-pass)filtered signal is returned.

Reduction: Only every n-th spectrum is returned. Use of a large TimeConstant can provide useful reduction of the data volume.

那你用MATLAB打开REMEZ.M程序看看remez函数是怎么遍写出来的．不过估计很难能看明白（我没打开，因为MATLAB被我卸了），为了适应这个函数的通用性，肯定在编写过程中加了N多的参量和情况分析，而且也肯定调用了不少其他相关的函数．一般自己写的话会比较困难．

个人建议你不要加这个算法到程序梁弊中，而是看看这个函数对你设计出的滤波器有哪些限制和影响．这样在分析滤波器特性中可以分析出仿真本身对设计的影响．一般用MATLAB做仿真分析能有合适的函数不会自己写，毕竟MATLAB仿真的核心是分析，而不是编程．

我也只能帮你到这里了，我没能力分析remez函数了．哗渣键看书上说用的是乱巧remez算法和切比雪夫近似来拟合期望的频率响应＠＠～

---