牛顿法.txt

约束最优化方法－－最速下降法(也叫梯度法)

syms x1 x2  xk xk2 gk gk2 Gk Gk2 g G i ni zhuan n fx
f=x1^2+x2^2+x1*x2;
g=jacobian(f);
G=jacobian(g);
x1=3;x2=-1;
gk=subs(g);
Gk=subs(G);
i=1
ni=inv(Gk);
n=20;
xk=[3,-1]
ni*gk'
while(n>0.0001)
xk=(xk'-ni*gk')'
x1=xk(1);
x2=xk(2);
gk=subs(g);
Gk=subs(G);
ni=inv(Gk);
n=norm(gk);
i=i+1
fx=subs(f)
end