shift model on experimental data with least squares method

Question

Stefan 2014-4-1

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/124035-shift-model-on-experimental-data-with-least-squares-method

回答： Star Strider 2014-10-22

Hi,

I have a set of experimental data (x,y) and another set of data from a model (xm,ym) which takes as input some experimental values. The output for the model is basically: data=model(input_unrelated_to_x); where plot(data.xm,data.ym) creates the curve of the model. Also, diff(x) and diff(xm) are almost identical.

I need to shift the model curve on the experimental curve on both x-axis and y-axis so that they overlap. Can it be done with the least squares method and how? Is there a better method ?

Thanks, Stefan

2 个评论
显示无隐藏无

Raj Raj 2014-9-30

Hi Stefan,

Were you able to solve this issue. I am currently stuck up with the same task.

Regards Raj

Stefan 2014-10-22

Hi Raj,

Unfortunately not, have you succeeded in solving this issue ? If not, I will try to ask again in a clearer way, I still need it.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Star Strider 2014-10-22

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/124035-shift-model-on-experimental-data-with-least-squares-method#answer_156172

It may be necessary for you to include the x- and y-shifts (offsets) in your model equation as separate parameters to be estimated. That would be my approach. I cannot be more specific without seeing your model.