Question

请教matlab高手，在多元非线性拟合当中出现如下问题，该如何解决？非常感谢！

x=[0,20,1200,160;61,20,1200,160;118,20,1200,160;169,20,1200,160;216,20,1200,160;261,20,1200,160;382,20,1200,160;450,20,1200,160;
0,40,1200,160;63,40,1200,160;123,40,1200,160;177,40,1200,160;227,40,1200,160;272,40,1200,160;316,40,1200,160;357,40,1200,160;450,40,1200,160;
0,60,1200,160;64,60,1200,160;128,60,1200,160;185,60,1200,160;236,60,1200,160;282,60,1200,160;327,60,1200,160;369,60,1200,160;
0,20,1400,180;60,20,1400,180;88,20,1400,180;122,20,1400,180;167,20,1400,180;237,20,1400,180;0,40,1400,180;63,40,1400,180;93,40,1400,180;150,40,1400,180;202,40,1400,180;250,40,1400,180;292,40,1400,180;323,40,1400,180;
0,60,1400,180;32,60,1400,180;68,60,1400,180;100,60,1400,180;128,60,1400,180;157,60,1400,180;210,60,1400,180;256,60,1400,180;279,60,1400,180;364,60,1400,180;
0,20,1600,200;60,20,1600,200;118,20,1600,200;169,20,1600,200;260,20,1600,200;303,20,1600,200;382,20,1600,200;0,40,1600,200;63,40,1600,200;123,40,1600,200;177,40,1600,200;226,40,1600,200;271,40,1600,200;315,40,1600,200;356,40,1600,200;
0,60,1600,200;64,60,1600,200;128,60,1600,200;184,60,1600,200;234,60,1600,200;280,60,1600,200;325,60,1600,200;443,60,1600,200;478,60,1600,200;669,60,1600,200];
y=[-20;-14;-12;-10;-4;-4;-2;0;-40;-30;-22;-16;-10;-8;-2;-1;0;-58;-44;-32;-28;-16;-10;-2;0;
-24;-18;-12;-6;-2;0;-48;-36;-30;-18;-12;-10;-6;-2;-66;-64;-56;-42;-34;-28;-14;-14;-4;-2;
-22;-18;-14;-8;-4;-2;-1;-38;-32;-26;-16;-12;-6;-4;-1;-62;-52;-38;-30;-22;-14;-10;-8;-4;0];
f=@(A,X)exp(A(1).*(X(:,2)./X(:,3)).^A(2).*(X(:,4)./600).^A(3).*(550./X(:,3)).^A(4).*(X(:,3)./(X(:,3)-X(:,2))).^A(5)).*exp(X(:,1)./600).*X(:,2);
[ahat,r,j]=nlinfit(x,y,f,[-1.877;0.063;0.441;0.989;7.591]),ci=nlparci(ahat,r,j)
Warning: Rank deficient, rank = 0, tol = 0.0000e+000.
> In nlinfit>LMfit at 312
In nlinfit at 166
Warning: Some columns of the Jacobian are effectively zero at the solution, indicating that the model is insensitive to some of its parameters. That may be because those parameters are
not present in the model, or otherwise do not affect the predicted values. It may also be due to numerical underflow in the model function, which can sometimes be avoided by choosing
better initial parameter values, or by rescaling or recentering. Parameter estimates may be unreliable.
> In nlinfit at 231

Answer 1

这样做拟合是不行的，你试试其他的算法啊

Answer 2

靠死……我头炸了楼主真牛