Non-linear curve fitting

13 次查看（过去 30 天）
Filip Jackowski
Filip Jackowski 2019-6-5
编辑： Matt J 2019-6-5
I apologize in advance if there there are similar cases of this question already getting answered, I tried following along but had trouble applying the given solutions to my own issue and would really appreciate a step-by-step solution.
I have a case where I've gathered experimental data, x and y where both x and y are vectors and im trying to fit it to the equation y = c1 + (c2*x)/(1 + c3*x)
I need to find a way to solve for c1, c2, and c3. Any help would be tremendously appreciated

请先登录，再回答此问题。

采纳的回答

Star Strider
Star Strider 2019-6-5
You need to define your model function as a function, then use some optimization function to fit the parameters.
Example:
yfcn = @(c1,c2,c3,x) c1 + (c2.*x)./(1 + c3.*x); % Model Function (Objective Function)
x = rand(10,1); % Create Data
y = rand(10,1); % Create Data
[B,fval] = fminsearch(@(b) norm(y - yfcn(b(1),b(2),b(3),x)), [1; 2; 1]); % Estimate Parameters
C1 = B(1)
C2 = B(2)
C3 = B(3)
There are many other functions you can use to estimate the parameters, including fmincon, fminunc, nlinfit, lsqcurvefit, and ga, as well as many others. See the documentation for the various functions (including norm here) to understand how to use them.

更多回答（1 个）

Matt J
Matt J 2019-6-5
编辑：Matt J 2019-6-5
fminspleas would be good for this (Download Here)
c3_guess = ____; %As good a guess as you know of c3
flist={1,@(c3,x)x./(1+c3.*x)};
[c3,c12]=fminspleas(flist,c3_guess);
c=[c12,c3];

类别

在 Help Center 和 File Exchange 中查找有关 Get Started with Curve Fitting Toolbox 的更多信息

标签

产品

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by