I would like Matlab to solve a system of equations iteratively, so that two among the three solutions of the (i-1)-th iteration are used as input data for the i-th iteration.
This is the code I'm trying to run,
clear vars close all clc
global Beta Tchnm Deltat Alpha Teta Deltaz mchn Tfm Gamma mf Delta Epsi Ta
Alpha = 1.61659e-07; Beta = 0.01731788; Gamma = 1.008413695; Delta = 1.213117234; Epsi = 0.001017152; Teta = 0.461538462; Deltaz = 0.001; Deltat = 0.005; mf = 0.03; mchn = 0.03; T0 = 89.3 + 273.15; Ta = 298.15;
solveoptions = optimoptions('fsolve','Display','iter',... 'Algorithm','levenberg-marquardt');
X0 = [350 350 350]; Sol = zeros(10,3);
for i = 1:10 if i == 1 Tchnm(i) = T0; Tfm(i) = T0; Sol(i,:) = fsolve(@(x) Equations(x),X0,solveoptions) else Tchnm(i) = Sol(i-1,2); Tfm(i) = Sol(i-1,3); Sol(i,:) = fsolve(@(x) Equations(x),X0,solveoptions) end end
function f = Equations(x)
global Beta Tchnm Deltat Alpha Teta Deltaz mchn Tfm Gamma mf Delta Epsi Ta
Tchnp = x(1); Tchn = x(2); Tf = x(3);
Tfmean = (Tf + Tfm)/2; Tchnmean = (Tchn + Tchnm)/2;
f(1) = -(Tchn - Tchnm)/Deltat + (Alpha*(Tchnp - 2*Tchn + Tchnm)/Deltaz^2) + Beta*(1 + mchn*Tchn); f(2) = -(Tf - Tfm)/Deltaz + Gamma*(1 + mf*Tfmean) - Delta*(Tfmean - Tchnmean) - Epsi*(Tfmean - Ta); f(3) = (Tchn - Tchnm) + Teta*(Tchn - Tf);
f = f'; end
This is the message I get: Subscripted assignment dimension mismatch. I think that the error is in the for loop.
Thanks in advance, Riccardo Rinaldi
