nonlinear system of equations

Hello all, I want to solve for three variables: epsilon=m(1), Jr(epsilon)=m(2), dr=m(3); The equations are quite complicated, as follows: Pd=0.015, Fr=8900, Z=9, Kn=3.5893e+005; Equation 1: epsilon= 0.5*(1-0.5*Pd/dr) Equation 2: dr=Pd/2 + (Fr/(Z*Kn*Jr(epsilon))^(1/1.5) Equation 3: Jr(epsilon)=(1/2pi)*{integration(-th1 to th1)[1-(0.5/epsilon)*(1-cos(theta))]^(1.5)]*cos(theta)}; where th1=cosine inverse (Pd/(2dr)) Actually theta is not the variable, its limit just depend on dr the third variable in question.

To solve for this system, I have written following code: function [ n ] = loaddistr( m ) Fr=8900; Pd=0.015; Z=9; Kn=3.5893e+005; n=[m(1)-0.5*(1-0.5*Pd/m(3)); m(3)-Pd/2+(Fr/(Z*Kn*m(2)))^(1/1.5); m(2)-(0.5/pi)*int((1-(1/(2*m(1)))*(1-cos(theta)))^(1.5)*cos(theta), theta, -acosd(0.5*Pd/m(3)), acosd(0.5*Pd/m(3)))]; %UNTITLED Summary of this function goes here % Detailed explanation goes here

end

when I try to run this as m0=[0.42, 0.21512, 0.062283] p=fsolve(@loaddistr, m0)

I am getting following error: ??? Undefined function or variable 'theta'.

Error in ==> loaddistr at 6 n=[m(1)-0.5*(1-0.5*Pd/m(3)); m(3)-Pd/2+(Fr/(Z*Kn*m(2)))^(1/1.5);

Error in ==> fsolve at 248 fuser = feval(funfcn{3},x,varargin{:});

Caused by: Failure in initial user-supplied objective function evaluation. FSOLVE cannot continue.

 Can someone please tell me where I am going wrong. Is my approach is correct?

Thanks in advance,

Nikhil

[Matt J]

Please highlight and format your code using the

toolbar button. Currently, it is not readable.

[Nikhil]

 function [ n ] = loaddistr( m ) 
Fr=8900; Pd=0.015; Z=9; Kn=3.5893e+005; 
n=[m(1)-0.5*(1-0.5*Pd/m(3)); 
m(3)-Pd/2+(Fr/(Z*Kn*m(2)))^(1/1.5); 
m(2)-(0.5/pi)*int((1-(1/(2*m(1)))*(1-cos(theta)))^(1.5)*cos(theta), theta, -acosd(0.5*Pd/m(3)), acosd(0.5*Pd/m(3)))]; 
end

[Nikhil]

 I want to solve for three variables: 
epsilon=m(1), Jr(epsilon)=m(2), dr=m(3); 
The equations are quite complicated, as follows:
 Pd=0.015, Fr=8900, Z=9, Kn=3.5893e+005; 
Equation 1: epsilon= 0.5*(1-0.5*Pd/dr)
 Equation 2: dr=Pd/2 + (Fr/(Z*Kn*Jr(epsilon))^(1/1.5)
 Equation 3: Jr(epsilon)=(1/2pi)*{integration(-th1 to th1)[1-(0.5/epsilon)*(1-cos(theta))]^(1.5)]*cos(theta)}; 
where th1=cosine inverse (Pd/(2dr))
 Actually theta is not the variable, its limit just depend on dr the third variable in question.

To solve for this system, I have written above code. Can someone please tell me where I am going wrong. Is my approach is correct? Thanks in advance,

[Mohammad Monfared]

I think you can define an anonymous function earlier in your main function, hence fun became like:

function [ n ] = loaddistr( m ) 
Fr=8900; Pd=0.015; Z=9; Kn=3.5893e+005; 
FUN=@(theta)(1-(1/(2*m(1)))*(1-cos(theta)))^(1.5)*cos(theta);
n=[m(1)-0.5*(1-0.5*Pd/m(3)); 
m(3)-Pd/2+(Fr/(Z*Kn*m(2)))^(1/1.5); 
m(2)-(0.5/pi)*integral(FUN, -acosd(0.5*Pd/m(3)), acosd(0.5*Pd/m(3)))];