Attempted to access c(4) ,,, lsqnonlin
조회 수: 2 (최근 30일)
이전 댓글 표시
Hi i'm getting two errors
- Attempted to access c(4); index out of bounds because numel(c)=3
- Error in lsqnonlin Caused by: Failure in initial user-supplied objective function evaluation. LSQNONLIN cannot continue.
Any suggestions on how I can solve there errors ?
function V2mainprogram
clc;
clear;
format long;
fid=fopen('referencerawdata.txt','rt');
c=textscan(fid,'%f %f %f',1);
Ro=cell2mat(c(1,1));
Ri=cell2mat(c(1,2));
theta0=cell2mat(c(1,3));
theta0=theta0.*pi./180; %converting to radians
c=textscan(fid,'%f %f %f %f','Headerlines',1);
fclose(fid);
Pexp=cell2mat(c(1));
riexp=cell2mat(c(2));
lambdaexp=cell2mat(c(3));
Fexp=cell2mat(c(4));
options=optimset('largescale','off','MaxFunEvals',1e100,'TolFun',1e-30,'TolX',1e-30,'MaxIter',5e3,'Algorithm','levenberg-marquardt','Display','off');
w=[0.4679139345726910 0.3607615730481386 0.1713244923791704 0.4679139345726910 0.3607615730481386 0.1713244923791704];
ksi=[-0.2386191860831969 -0.6612093864662645 -0.9324695142031521 0.2386191860831969 0.6612093864662645 0.9324695142031521];
model='G';
iniconstants=[0.1 0.1 0.1];
RadLambexp=[riexp lambdaexp];
constants=lsqnonlin(@(x) errPF(x,RadLambexp,Pexp,Fexp,Ro,Ri,theta0,ksi,w,model),iniconstants,[],[],options);
fprintf('\n');
fprintf('G model constants\n');
fprintf('c1=%f\n',constants(1));
fprintf('c2=%f\n',constants(2));
fprintf('c3=%f\n',constants(3));
iniguess=[riexp lambdaexp];
RadLamb=lsqnonlin(@(x) errPF(constants,x,Pexp,Fexp,Ro,Ri,theta0,ksi,w,model),iniguess,[],[],options);
[Pressure,Force]=PF(constants,RadLamb,Ro,Ri,theta0,ksi,w,model); %uses optimized values to get theoretical pressure
RadLambG=RadLamb;
PressureG=Pressure;
ForceG=Force;
figure(1);
subplot(2,1,1);
plot(Pexp.*1000,riexp,('-r')); hold on;
plot(PressureFung.*1000,RadLambFung(:,1),'-k',PressureTH.*1000,RadLambTH(:,1),'--',PressureG.*1000,RadLambG(:,1),'--');
ylabel ('Inner radius [mm]');
xlabel('Inner pressure [KPa]');
subplot(2,1,2);
plot(Pexp.*1000,riexp,('-r')); hold on;
plot(PressureFung.*1000,RadLambFung(:,2),'-k',PressureTH.*1000,RadLambTH(:,2),'--',PressureG.*1000,RadLambG(:,2),'--');
ylabel ('Longitudinal stretch ratio');
xlabel('Inner pressure [KPa]');
end
%*****************************
function error=errPF(c,RadLamb,Pexp,Fexp,Ro,Ri,theta0,ksi,w,model)
[Pt,Ft]=PF(c,RadLamb,Ro,Ri,theta0,ksi,w,model);
errp=(Pt-Pexp).*Ri^2;
errf=(Ft-Fexp);
error=[errp;errf];
end
%***********************************
function[Pt,Ft]=PF(c,RadLamb,Ro,Ri,theta0,ksi,w,model)
ri=RadLamb(:,1);
lambda=RadLamb(:,2);
P=zeros(length(ri),1);
F=zeros(length(ri),1);
ro=sqrt(((Ro.^2-Ri.^2).*theta0)./(pi.*lambda)+ri.^2);
for i=1:length(ksi)
r=((ro+ri)./2+((ro-ri)./2).*ksi(i));%gd
R=sqrt(Ri.^2+(((pi.*lambda)./theta0).*(r.^2-ri.^2)));%gd
Ezz=0.5*((lambda.^2)-1);%gd
Ett=0.5*((((pi.*r)./(theta0.*R)).^2)-1);%gd
calcPartials=str2func(model);
[dWtt,dWzz]=calcPartials(c,Ett,Ezz);
P=P+w(i).*(((((pi.*r)./(theta0.*R)).^2).*dWtt)./r);%gd
F=F+w(i).*((2.*(lambda.^2).*dWzz)-(((pi.*r)./(theta0.*R)).^2).*dWtt).*r;%gd
end
Pt=((ro-ri)./2).*P;
Ft=pi.*((ro-ri)./2).*F;
end
%***************************************
function[dWtt,dWzz]=G(c,Ett,Ezz)
dWtt=1/2.*c(1).*((2.*c(2).*Ett)-(c(3).*(1./(((2.*Ett+1).^2).*(2.*Ezz+1)))).*(1./((2.*Ett+1).*(2.*Ezz+1)))).*(exp((c(2).*(Ett.^2))+(c(3).*(Ezz.^2))+((c(4)).*(1/4).*(((1./((2.*Ett+1).*(2.*Ezz+1))))-1.^2))));
dWzz=1/2.*c(1).*((2.*c(3).*Ezz)-(c(3).*(1./(((2.*Ezz+1).^2).*(2.*Ett+1)))).*(1./((2.*Ett+1).*(2.*Ezz+1)))).*(exp((c(2).*(Ett.^2))+(c(3).*(Ezz.^2))+((c(4)).*(1/4).*(((1./((2.*Ett+1).*(2.*Ezz+1))))-1.^2))));
end
댓글 수: 0
채택된 답변
Alan Weiss
2014년 3월 26일
Well, I see in
function[dWtt,dWzz]=G(c,Ett,Ezz)
lines with
...+((c(4)).*...
If c has only three components, then these lines will definitely thow an error.
Alan Weiss
MATLAB mathematical toolbox documentation
추가 답변 (0개)
참고 항목
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!