Find the steady states of a system.

2 ビュー (過去 30 日間)
Ciaran
Ciaran 2015 年 1 月 7 日
回答済み: Ingrid Tigges 2015 年 1 月 7 日
I am building a SimBiology Model. I want to simulate from randomly generated initial conditions numerous times and find the steady state of a particular specie (PLS). MY current code does this in a graphical sense but I need the actual numbers in a vector (for example) for subsequent use. Does anybody know how to do this? My current code is:
%Random Number Generations and Assignment
No_of_simulations = 10; %i.e. points
number_of_parameters = 4; %i.e. dimenions
lb=[1e-3];
ub=[0.1];
x = lhsdesign(number_of_parameters ,No_of_simulations);
D = bsxfun(@plus,lb,bsxfun(@times,x,(ub-lb)));
for i=x,
n1 = i(1);
n2 = i(2);
n3 = i(3);
n4 = i(4);
%Model generation.
Mobj = sbiomodel('u-PA_parameter_generation');
comp_obj = addcompartment(Mobj, 'plasma');
%Reaction 1
Robj1 = addreaction(Mobj, 'Pro-u-PA + PLG -> PLS + Pro-u-PA');
Kobj1= addkineticlaw(Robj1, 'MassAction');
Pobj1 = addparameter(Kobj1, 'keff_zymogen',0.035);
set(Kobj1, 'ParameterVariableNames','keff_zymogen');
%Reaction 2
Robj2 = addreaction(Mobj, 'PLS + Pro-u-PA -> PLS + u-PA');
Kobj2= addkineticlaw(Robj2, 'MassAction');
Pobj2 = addparameter(Kobj2, 'keff_PLS',40);
set(Kobj2, 'ParameterVariableNames','keff_PLS');
%Reaction 3
Robj3 = addreaction(Mobj, 'u-PA + PLG -> u-PA + PLS');
Kobj3= addkineticlaw(Robj3, 'MassAction');
Pobj3 = addparameter(Kobj3, 'keff_pos',0.9);
set(Kobj3, 'ParameterVariableNames','keff_pos');
%Reaction 4
Robj4 = addreaction(Mobj, 'Pro-u-PA -> null');
Kobj4= addkineticlaw(Robj4, 'MassAction');
Pobj4 = addparameter(Kobj4, 'u1',0.084);
set(Kobj4, 'ParameterVariableNames','u1');
%Reaction 5
Robj5 = addreaction(Mobj, 'PLG -> null');
Kobj5= addkineticlaw(Robj5, 'MassAction');
Pobj5 = addparameter(Kobj5, 'u2',0.032);
set(Kobj5, 'ParameterVariableNames','u2');
%Reaction 6
Robj6 = addreaction(Mobj, 'PLS -> null');
Kobj6= addkineticlaw(Robj6, 'MassAction');
Pobj6 = addparameter(Kobj6, 'u1',0.084); %Same as R4
set(Kobj6, 'ParameterVariableNames','u1');
%Reaction 7
Robj7 = addreaction(Mobj, 'u-PA -> null');
Kobj7= addkineticlaw(Robj7, 'MassAction');
Pobj7 = addparameter(Kobj7, 'u1',0.084); %Same as R4
set(Kobj7, 'ParameterVariableNames','u1');
%Reaction 8
Robj8 = addreaction(Mobj, 'null -> Pro-u-PA');
Kobj8= addkineticlaw(Robj8, 'MassAction');
Pobj8 = addparameter(Kobj8, 'a1',0.0032);
set(Kobj8, 'ParameterVariableNames','a1');
%Reaction 9
Robj9 = addreaction(Mobj, 'null -> PLG');
Kobj9= addkineticlaw(Robj9, 'MassAction');
Pobj9 = addparameter(Kobj9, 'a2',0.01);
set(Kobj9, 'ParameterVariableNames','a2');
%setting species concentrations
Sobj1 = sbioselect(Mobj,'Type','species','Name','Pro-u-PA');
set(Sobj1, 'InitialAmount',n1)
Sobj2 = sbioselect(Mobj,'Type','species','Name','PLG');
set(Sobj2, 'InitialAmount',n2)
Sobj3 = sbioselect(Mobj,'Type','species','Name','PLS');
set(Sobj3, 'InitialAmount',n3)
Sobj4 = sbioselect(Mobj,'Type','species','Name','u-PA');
set(Sobj4, 'InitialAmount',n4)
%simulate
config = getconfigset(Mobj);
set(config,'StopTime',1000)
[t_ode, x_ode, names] = sbiosimulate(Mobj);
hold on
figure;
set(gcf, 'color','white');
plot(t_ode, x_ode(:,1:end));
legend(names)
end
Thanks in Advance

採用された回答

Ciaran
Ciaran 2015 年 1 月 7 日
Basically the answer was to initialize the figure before the for loop

その他の回答 (1 件)

Ingrid Tigges
Ingrid Tigges 2015 年 1 月 7 日
Given that the steady state is defined as dx/dt=0 which is approximately delta x/delta t you can check whether the difference in x of two consecutive time points is 0
diff_x= diff(x_ode);
Due to numeric errors you want to have those differences that are below a certain threshold:
threshold = 1e-15;
x_equals_0 = diff_x<threshold;
The indices of those values in x_equals_0 that are 1 belong to points where dx/dt=0.

コミュニティ

その他の回答  SimBiology コミュニティ

カテゴリ

Help Center および File ExchangeExtend Modeling Environment 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!

Translated by