X = [R1,Xt_1,Xt_2,Xt_3];
Y = y;
modelfun = @(1+exp(b(1)*(b(2)-b(3))*X(:,4)*(X(:,2)-X(:,1)*X(:,3)))).^(-1).*b(2)*X(:,2)+(1-(1+exp(b(1)*(b(2)-b(3))*X(:,4)*(X(:,2)-X(:,1)*X(:,3)))).^(-1)).*b(3)*X(:,2);
beta0=[-10, 0.5, 0.5];
mdl = NonLinearModel.fit(X,Y,modelfun,beta0);

2楼: Originally posted by dingd at 2015-03-06 09:42:43
"NLLS"是否就是指"NonLinear Least Squares"？即非线性最小二乘法，试试1stOpt吧，不需要初值，简单易懂。