Smoothing a noisy signal

Question

David Polcari 2012-6-13

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/41015-smoothing-a-noisy-signal

Hey everyone,

I need some help smoothing out a noisy signal. I know of one or two methods but I am not sure what the best way of doing it is. Can someone tell me what is the most efficient way of smoothing?

Thanks for the help!

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

David Polcari 2012-6-13

Here is my dataset:

voltage_B1 = [1.0110

1.0110

1.0134

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0098

1.0134

1.0122

1.0134

1.0122

1.0110

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0122

1.0110

1.0122

1.0134

1.0122

1.0134

1.0147

1.0122

1.0134

1.0110

1.0122

1.0147

1.0134

1.0147

1.0134

1.0147

1.0134

1.0122

1.0134

1.0147

1.0159

1.0134

1.0147

1.0134

1.0147

1.0134

1.0147

1.0122

1.0134

1.0147

1.0159

1.0147

1.0134

1.0159

1.0147

1.0134

1.0122

1.0134

1.0147

1.0122

1.0134

1.0159

1.0122

1.0098

1.0134

1.0122

1.0098

1.0122

1.0098

1.0110

1.0122

1.0098

1.0122

1.0110

1.0122

1.0098

1.0134

1.0110

1.0122

1.0110

1.0122

1.0098

1.0110

1.0122

1.0110

1.0122

1.0134

1.0122

1.0110

1.0122

1.0110

1.0122

1.0134

1.0159

1.0110

1.0134

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0098

1.0134

1.0122

1.0134

1.0122

1.0110

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0122

1.0110

1.0122

1.0134

1.0122

1.0134

1.0147

1.0122

1.0134

1.0110

1.0122

1.0147

1.0134

1.0147

1.0134

1.0147

1.0134

1.0122

1.0134

1.0147

1.0159

1.0134

1.0147

1.0134

1.0147

1.0134

1.0147

1.0122

1.0134

1.0147

1.0159

1.0147

1.0134

1.0159

1.0147

1.0134

1.0122

1.0134

1.0147

1.0122

1.0134

1.0159

1.0122

1.0098

1.0134

1.0122

1.0098

1.0122

1.0098

1.0110

1.0122

1.0098

1.0122

1.0110

1.0122

1.0098

1.0134

1.0110

1.0122

1.0110

1.0122

1.0098

1.0110

1.0122

1.0110

1.0122

1.0134

1.0122

1.0110

1.0122

1.0110

1.0122

1.0134

1.0159

1.0122

1.0134

1.0147

1.0134

1.0147

1.0122

1.0147

1.0134

1.0122

1.0147

1.0134

1.0122

1.0134

1.0122

1.0147

1.0134

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0147

1.0134

1.0122

1.0134

1.0147

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0171

1.0134

1.0122

1.0110

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0110

1.0147

1.0159

1.0122

1.0134

1.0122

1.0159

1.0147

1.0159

1.0147

1.0134

1.0159

1.0122

1.0147

1.0110

1.0134

1.0147

1.0134

1.0147

1.0134

1.0159

1.0122

1.0147

1.0134

1.0147

1.0134

1.0147

1.0122

1.0159

1.0147

1.0134

1.0147

1.0159

1.0147

1.0159

1.0147

1.0159

1.0147

1.0159

1.0147

1.0159

1.0171

1.0134

1.0159

1.0134

1.0159

1.0147

1.0159

1.0147

1.0159

1.0171

1.0159

1.0171

1.0159

1.0147

1.0159

1.0134

1.0171

1.0159

1.0171

1.0159

1.0171

1.0147

1.0171

1.0147

1.0171

1.0159

1.0183

1.0195

1.0171

1.0183

1.0134

1.0171

1.0134

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0171

1.0134

1.0122

1.0110

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0122

1.0134

1.0110

1.0147

1.0159

1.0122

1.0134

1.0122

1.0159

1.0147

1.0159

1.0147

1.0134

1.0159

1.0122

1.0147];

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Walter Roberson 2012-6-15

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/41015-smoothing-a-noisy-signal#answer_50806

The most efficient way of smoothing is to leave the data untouched.

conv(signal, ones(1,3)/3) is fairly efficient computationally, but you could probably do better if you created a custom mex routine.

If you were instead perhaps wondering about the most effective way of smoothing, then you need to define your goals for the results.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

Answer 2

Sandarva Khanal 2012-6-15

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/41015-smoothing-a-noisy-signal#answer_50818

Like Walter said, you need to define your goals for the results. One simple way can be to use a running average of may be about 3 - 5 data from your data set.

For example, if you are interested in finding the running average of 3 data-points from your data set, your 2nd data could be the average of first three data-points. Your 3rd data will be the average of data-point indexed 2 to 4, and so on....remember that in this case, you will not usually change the first data-point.