Parameter estimation using Nonlinear Regression

Hi all

I'm trying to estimate parameters of a function by nonlinear regression using "NonLinearModel.fit". My code is as follows. I have some problems. First of all all the pvalues and confidence intervals are NAN (I computed the jacobian matrix and it's not zero or inf...) and also it does not improve the coefficients and the estimated parameters are exactly the initial values which I provide!!

Thank you for your help in advance.

clc

clear all

[data] = xlsread('Book1.xlsx');

X=data(:,1);

y=data(:,2);

modelfun = @(b,x) b(1) .* exp(-b(2).*(exp(-b(3)./x(:,1)).*1800));

beta0 = [20 400 120];

opts = statset('Display','iter','TolFun',1e-20);

mdl = NonLinearModel.fit(X,y,modelfun,beta0,'Options',opts);

alpha = 0.05; % this is for 95% confidence intervals

ci = coefCI(mdl,alpha);