Help for Monte Carlo method

Hi guys . i am using monte carlo method to calculate uncertainity. My matlab codes are working properly-However i want to ask question . I wiil be very glad if i recieve an answer. Lets assume that one of my parameter has two uncertainty art like(Type a and Type B errors) Do you know how can i add both of them in my model. exp: F=m.a mass has 2 uncertainty type. Assume that first one Type A normal distribution and second one Type B rectangular. how can i add them

mass_rc=random('norm',100000,0.05,1,M);

Answer 1

[Torsten]

If the errors are independent and you know a formula how they combine (Uc=f(Ua,Ub)), you can independently generate Ua and Ub from their distributions, calculate Uc according to your combination formula and add/subtract (with probability 1/2) Uc from the mean value of the variable in question.

But the function f is not Monto-Carlo specific - you must know it in advance.

Best wishes

Torsten.

[cemsi888]

I think i could not describe my problem.. example will help us... F=m.a (basic formula) mass has two error components. Type A and Type B (Device error). In order to calculate overall uncertainty of F i have to calculate uncertainty of m and a. and assume that accelaration has not any type B error. is it possible that to calculate at the same time overall uncertainty like gauss methode

[cemsi888]

   mass_rc=random('norm',100000,0.05,1,M);
    dmass_rc=random('norm',1.234,0.02,1,M);
    air=random('unif',1.1,1.3,1,M);
    air0=1.2;
    dens_r=random('unif',7950,8050,1,M);
    dens_w=random('unif',7000,9000,1,M);
    non_mass=100000;
    y=(mass_rc+dmass_rc).*[1+(air-air0).*(1./dens_w-1./dens_r)]-non_mass;
for example these are my codes and assume that mass_rc has type B error. have can i add it ???