Trouble solving ODE equations
古いコメントを表示
I am trying to solve these differential equations in order to plot graphs, however, when running the code I am receiving multiple errors but I'm unsure why as I can't see why the code is not running smoothly? I'm looking for some insight on why this might be happening and how I can solve the problem?
Equations and Parameter values:
function dydt = Dunster(t,y)
%Parameters and their values%
K1a = 20;
Y1a = 1;
K1b = 0.25;
K2a = 14;
K2am = 72;
K2b = 2.6;
K2bm = 10;
K2c = 24;
K2cm = 20;
K3a = 10;
K3b = 0.05;
K3c = 24;
K3cm = 20;
K4a = 2.3;
K4am = 58;
K4b = 2000;
K4bm = 210;
K4c = 1.3;
K5a = 0.0014;
K5b = 0.35;
K6 = 2000;
K6m = 2000;
Kx = 50;
Kb = 0.01;
Lb = 1;
Lx = 1;
Bxva = 0;
H = K1a*Y1a*exp(-Y1a*t);
bl = 0.5*((Kb+Lb+Bxva)-sqrt((Kb+Lb+Bxva)^2 - 4*Lb*Bxva));
%Equation 1%
%Generation and inactivation of factor Xa%
F1= y(1);
%y(1)= -K6*F1;
% Factor I
%Equation 2%
F5=y(2);
%y(2)=(-K2a*F2a*F5)/(F5+K2am(1+F1/K6m) - ((K2b*F10a*F5)/F5+K2bm(1+F2/K4am)));
%Equation 3%
F2=y(3);
%y(3)= ((-K4a*F1*F2)/F2+K4am(1+F5/K2bm))-((K4b*b1*F2)/(F2+K4bm));
%Equation 4%
C=y(4);
%y(4)= (-K5a*C);
%Equation 5%
F10a=y(5);
%y(5)= H +(K3c*Ca*Bxva)/(Bxva+K3cm) - K1b*F10a - K3a*F10a*F5a;
%Equation 6%
F5a=y(6);
%y(6)= (K2a*F2a*F5)/(F5+K2am(1+F1/K6m)) + (K2b*F10a*F5a)/(F5+K2bm(1+F1/K4am))+ K3b*Bxva - (K2c*Ca*F5a)/(F5a+K2am) - K3a*F10a*F5a;
%Equation 7%
Bxva=y(7);
%y(7)= K3a*F10a*F5a - K3b*Bxva - (K3c*Ca*Bxva)/(Bxva+K3cm);
%Equation 8%
F2a=y(8);
%y(8)= (K4a*F1*F2)/(F2+K4am(1+F5/K2bm)) + (K4b*BF2)/(F2+K4bm) - K4c*F2a;
%Equation 9%
Ca=y(9);
%y(9)= K5a*C - K5b*Ca;
%Equation 10%
F10i=y(10);
%y(10)= K1b*F10a + K3b*Bxva;
%Equation 11%
F5i=y(11);
%y(11) = (K2c*Ca*F5a)/(F5a+K2cm) + (K3c*Ca*Bxva/(Bxva+K3cm));
%Equation 12%
F2i=y(12);
%y(12) = K4c*F2a;
%Equation 13%
Ci=y(13);
%y(13) = K5b*Ca;
%Equation 14%
F1a=y(14);
%y(14) = K6*F1;
dydt=[
-K6*F1; % d F1 / dt
((-K2a*F2a*F5)/(F5+K2am*(1+F1/K6m)) - ((K2b*F10a*F5)/F5+K2bm*(1+F2/K4am))); % d F5 / dt
((-K4a*F1*F2)/(F2+K4am*(1+F5/K2bm))-((K4b*bl*F2)/(F2+K4bm))); % d F2 / dt
-K5a*C; % d C / dt
H+(K3c*Ca*Bxva)/(Bxva+K3cm) - K1b*F10a - K3a*F10a*F5a ;
% d F10a / dt
(K2a*F2a*F5)/(F5+K2am(1+F1/K6m)) + (K2b*F10a*F5a)/(F5+K2bm(1+F1/K4am))
+ K3b*Bxva - (K2c*Ca*F5a)/(F5a+K2am) - K3a*F10a*F5a; % d F5a / dt
K3a*F10a*F5a - K3b*Bxva - (K3c*Ca*Bxva)/(Bxva+K3cm); % d Bxva / dt
(K4a*F1*F2)/(F2+K4am(1+F5/K2bm)) + (K4b*BF2)/(F2+K4bm) - K4c*F2a ;
% d F2a / dt
K5a*C - K5b*Ca ; % d Ca / dt
K1b*F10a + K3b*Bxva ; % d F10i / dt
((K2c*Ca*F5a)/(F5a+K2cm) + (K3c*Ca*Bxva/(Bxva+K3cm))) ; % d F5i / dt
K4c*F2a ; % d F2i / dt
K5b*Ca ; % d Ci / dt
K6*F1 ; % d F1a / dt
];
end
Running Part of the Code:
tspan=[0 60];
F1=10000;
F5=30;
F2=1000;
C=100;
F10a=0;
F5a=0;
Bxva=0;
F2a=0;
Ca=0;
F10i=0;
F5i=0;
F2i=0;
Ci=0;
F1a=0;
y0=[F1;F5;F2;C;F10a;F5a;Bxva;F2a;Ca;F10i;F5i;F2i;Ci;F1a];
[t,y]=ode45(@Dunster, tspan, y0);%,[], pars);
plot(t/60, y(:,14)*1000, "r")
xlabel("Time [Min]")
ylabel("Concentration [nM]")
1 件のコメント
Rena Berman
2021 年 5 月 6 日
(Answers Dev) Restored edit
回答 (1 件)
Star Strider
2021 年 4 月 11 日
There were several missing multiplication operators, and a reference to ‘BF2’ that I corrected to ‘F2’, since there is no ‘B’ that I can find, so no missing multiplication operator there. With those edits, and concatenating an additional 0 to ‘y0’ to make its length equal to the number of differential equations, this runs without error:
function dydt = Dunster(t,y)
%Parameters and their values%
K1a = 20;
Y1a = 1;
K1b = 0.25;
K2a = 14;
K2am = 72;
K2b = 2.6;
K2bm = 10;
K2c = 24;
K2cm = 20;
K3a = 10;
K3b = 0.05;
K3c = 24;
K3cm = 20;
K4a = 2.3;
K4am = 58;
K4b = 2000;
K4bm = 210;
K4c = 1.3;
K5a = 0.0014;
K5b = 0.35;
K6 = 2000;
K6m = 2000;
Kx = 50;
Kb = 0.01;
Lb = 1;
Lx = 1;
Bxva = 0;
H = K1a*Y1a*exp(-Y1a*t);
bl = 0.5*((Kb+Lb+Bxva)-sqrt((Kb+Lb+Bxva)^2 - 4*Lb*Bxva));
%Equation 1%
%Generation and inactivation of factor Xa%
F1= y(1);
%y(1)= -K6*F1;
% Factor I
%Equation 2%
F5=y(2);
%y(2)=(-K2a*F2a*F5)/(F5+K2am(1+F1/K6m) - ((K2b*F10a*F5)/F5+K2bm(1+F2/K4am)));
%Equation 3%
F2=y(3);
%y(3)= ((-K4a*F1*F2)/F2+K4am(1+F5/K2bm))-((K4b*b1*F2)/(F2+K4bm));
%Equation 4%
C=y(4);
%y(4)= (-K5a*C);
%Equation 5%
F10a=y(5);
%y(5)= H +(K3c*Ca*Bxva)/(Bxva+K3cm) - K1b*F10a - K3a*F10a*F5a;
%Equation 6%
F5a=y(6);
%y(6)= (K2a*F2a*F5)/(F5+K2am(1+F1/K6m)) + (K2b*F10a*F5a)/(F5+K2bm(1+F1/K4am))+ K3b*Bxva - (K2c*Ca*F5a)/(F5a+K2am) - K3a*F10a*F5a;
%Equation 7%
Bxva=y(7);
%y(7)= K3a*F10a*F5a - K3b*Bxva - (K3c*Ca*Bxva)/(Bxva+K3cm);
%Equation 8%
F2a=y(8);
%y(8)= (K4a*F1*F2)/(F2+K4am(1+F5/K2bm)) + (K4b*BF2)/(F2+K4bm) - K4c*F2a;
%Equation 9%
Ca=y(9);
%y(9)= K5a*C - K5b*Ca;
%Equation 10%
F10i=y(10);
%y(10)= K1b*F10a + K3b*Bxva;
%Equation 11%
F5i=y(11);
%y(11) = (K2c*Ca*F5a)/(F5a+K2cm) + (K3c*Ca*Bxva/(Bxva+K3cm));
%Equation 12%
F2i=y(12);
%y(12) = K4c*F2a;
%Equation 13%
Ci=y(13);
%y(13) = K5b*Ca;
%Equation 14%
F1a=y(14);
%y(14) = K6*F1;
dydt=[-K6*F1; % d F1 / dt
((-K2a*F2a*F5)/(F5+K2am*(1+F1/K6m)) - ((K2b*F10a*F5)/F5+K2bm*(1+F2/K4am))); % d F5 / dt
((-K4a*F1*F2)/(F2+K4am*(1+F5/K2bm))-((K4b*bl*F2)/(F2+K4bm))); % d F2 / dt
-K5a*C; % d C / dt
H+(K3c*Ca*Bxva)/(Bxva+K3cm) - K1b*F10a - K3a*F10a*F5a ;
% d F10a / dt
(K2a*F2a*F5)/(F5+K2am*(1+F1/K6m)) + (K2b*F10a*F5a)/(F5+K2bm*(1+F1/K4am))
+ K3b*Bxva - (K2c*Ca*F5a)/(F5a+K2am) - K3a*F10a*F5a; % d F5a / dt
K3a*F10a*F5a - K3b*Bxva - (K3c*Ca*Bxva)/(Bxva+K3cm); % d Bxva / dt
(K4a*F1*F2)/(F2+K4am*(1+F5/K2bm)) + (K4b*F2)/(F2+K4bm) - K4c*F2a ;
% d F2a / dt
K5a*C - K5b*Ca ; % d Ca / dt
K1b*F10a + K3b*Bxva ; % d F10i / dt
((K2c*Ca*F5a)/(F5a+K2cm) + (K3c*Ca*Bxva/(Bxva+K3cm))) ; % d F5i / dt
K4c*F2a ; % d F2i / dt
K5b*Ca ; % d Ci / dt
K6*F1 ; % d F1a / dt
];
end
% % Running Part of the Code:
tspan=[0 60];
F1=10000;
F5=30;
F2=1000;
C=100;
F10a=0;
F5a=0;
Bxva=0;
F2a=0;
Ca=0;
F10i=0;
F5i=0;
F2i=0;
Ci=0;
F1a=0;
y0=[F1;F5;F2;C;F10a;F5a;Bxva;F2a;Ca;F10i;F5i;F2i;Ci;F1a;0];
[t,y]=ode45(@Dunster, tspan, y0);%,[], pars);
plot(t/60, y(:,14)*1000, "r")
xlabel("Time [Min]")
ylabel("Concentration [nM]")
.
6 件のコメント
Pr0t0nZ
2021 年 4 月 11 日
Star Strider
2021 年 4 月 11 日
My pleasure!
Put this after the ode45 call:
figure
for k = 1:size(y,2)
subplot(8,2,k)
plot(t/60, y(:,k)*1000, "r")
xlabel("Time [Min]")
ylabel("Concentration [nM]")
title(sprintf('y_{%d}',k))
end
pos = get(gcf, 'Position');
set(gcf, 'Position',pos+[0 -500 0 500])
The rest of the code is unchanged.
There are other options (such as plotting them all on the same axes), however with 15 variables with significantly different magnitudes, that would be difficult to interpret.
Experiment to get the result you want.
Pr0t0nZ
2021 年 4 月 11 日
Star Strider
2021 年 4 月 11 日
My pleasure!
Fro the plots:
for k = 1:size(y,2)
figure
plot(t/60, y(:,k)*1000, "r")
xlabel("Time [Min]")
ylabel("Concentration [nM]")
title(sprintf('y_{%d}',k))
set(gca, 'XScale','log') % Optional
end
Also consider:
y0=[F1;F5;F2;C;F10a;F5a;Bxva;F2a;Ca;F10i;F5i;F2i;Ci;F1a;0]+eps; % Adding ‘eps’ Optional, Prevents Some Variables From Being Identically 0
.
Pr0t0nZ
2021 年 4 月 11 日
Star Strider
2021 年 4 月 11 日
My pleasure!
If my Answer helped you solve your problem, please Accept it!
.
カテゴリ
ヘルプ センター および File Exchange で Programming についてさらに検索
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!