solving 5 non linear simultaneous equation
1 view (last 30 days)
Show older comments
Bibek Tripathi
on 30 Apr 2016
I have a set of 5 non-linear equations. first 4 equations are h,k,a,b --> in terms of variable 'bsfc'. the fifth equation bsfc^2 in terms of h,k,a,b. So i have 5 equations and 5 variables. This programming is basically finding the radius of the rotated ellipse. I have my x and y values too. However I am not getting the right solution.
The right solution is bsfc=246
I have my code here:
syms bsfc h k a b
y=86;
x=1717;
phi=0;
eqn1= h == 9.514*bsfc;
eqn2= k == 0.0001*(bsfc+1)^2-0.07*bsfc+19;
eqn3= a == 10* atan(bsfc/30 - 5.8) -3.5;
eqn4= b == 0.025 * atan(bsfc/20 - 10.3) -0.022;
eqn5= bsfc^2 == (((x-h)*cos(phi)+(y-k)*sin(phi))^2/(a^2)+ ((x-h)*sin(phi)-(y-k)*cos(phi))^2/(b^2))
sol=solve([eqn1,eqn2,eqn3,eqn4,eqn5],bsfc,h,k,a,b);
hSol = sol.h
kSol = sol.k
aSol = sol.a
bSol = sol.b
bsfcSol = sol.bsfc
0 Comments
Accepted Answer
Walter Roberson
on 1 May 2016
Given those equations, bsfc approximately 1154.643046927194 or approximately 1325.553297436922
Your proposed value 246 does not come at all close to satisfying eqn5
8 Comments
Walter Roberson
on 2 May 2016
syms bsfc h k a b x y phi;
Q = @(v) sym(v);
eqn1 = h == Q(9.514)*bsfc;
eqn2 = k == Q(0.1e-3)*(bsfc+1)^2-Q(0.7e-1)*bsfc+19;
eqn3 = a == 10*atan((1/30)*bsfc-Q(5.8))-Q(3.5);
eqn4 = b == Q(0.25e-1)*atan((1/20)*bsfc-Q(10.3))-Q(0.22e-1);
eqn5 = bsfc^2 == ((x-h)*cos(phi)+(y-k)*sin(phi))^2/a^2+((x-h)*sin(phi)-(y-k)*cos(phi))^2/b^2;
sol_hkab = solve([eqn2,eqn3,eqn4,eqn5],[h, k, a, b]);
eqn_bsfc1 = subs(eqn1, [h k a b], [sol_hkab.h(1), sol_hkab.k(1), sol_hkab.a(1), sol_hkab.b(1)]);
eqn_bsfc2 = subs(eqn1, [h k a b], [sol_hkab.h(2), sol_hkab.k(2), sol_hkab.a(2), sol_hkab.b(2)]);
xyt = [Q(1256.97), Q(7.076), Q(0.00044)];
bs = linspace(0,750,2000);
s_bsfc1 = subs(eqn_bsfc1, [x y phi], xyt);
eq1f = matlabFunction(feval(symengine,'lhs',s_bsfc1)-feval(symengine,'rhs',s_bsfc1));
eq1fr = @(x) real(eq1f(x));
eq1fi = @(x) imag(eq1f(x));
s_bsfc2 = subs(eqn_bsfc2, [x y phi], xyt);
eq2f = matlabFunction(feval(symengine,'lhs',s_bsfc2)-feval(symengine,'rhs',s_bsfc2));
eq2fr = @(x) real(eq2f(x));
eq2fi = @(x) imag(eq2f(x));
plot(bs, eq1fi(bs), 'b--', bs, eq2fi(bs), 'g--', bs, eq1fr(bs), 'b-', bs, eq2fr(bs), 'g-');
legend({'imag sol1', 'imag sol2', 'real sol1', 'real sol2'});
and then start zooming in on places where the imaginary is 0, looking for real solutions.
More Answers (2)
Alex Sha
on 31 Jul 2019
Look this result:
h: -4113.26496133792
bsfc: -432.338129213572
k: 67.8689272162873
a: -18.7135929756326
b: -0.0604868803921316
0 Comments
Alex Sha
on 1 Aug 2019
one more solution:
h: 10985.2739483926
bsfc: 1154.64304691955
k: 71.7260719049847
a: 11.9021369375982
b: 0.0167429176020812
0 Comments
See Also
Categories
Find more on Systems of Nonlinear Equations in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!