Fmincon Errors with objective function definition

조회 수: 9 (최근 30일)
Onur Gurler
Onur Gurler 2020년 12월 2일
댓글: Onur Gurler 2020년 12월 2일
Hi All,
I am trying to solve a simple fmincon problem and keep hitting the same error message:
Error using funhw2q1
Too many input arguments.
Error in fmincon (line 552)
initVals.f = feval(funfcn{3},X,varargin{:});
Error in HW2Q1solution (line 13)
xopt(i) = fmincon('funhw2q1', x0, [], [], [], [], LB, UB, 'const_hw2q1', [], j);
Caused by:
Failure in initial objective function evaluation. FMINCON cannot continue.
--------------------------------------------------------------------------------------------------------
Here is my main function:
x0 = [0 0];
LB = [-5 -5];
UB = [7 7];
i = 1;
for j = 0:0.01:1
xopt(i) = fmincon('funhw2q1', x0, [], [], [], [], LB, UB, 'const_hw2q1', [], j);
solution(j) = x(1)^2 + 10 * x(2) - 3 * X(1) * X(2);
i = i + 1;
end
-------------------------------------------------------------------------------------------------------
here is 'funhw2q1':
function f = funhw2q1(x)
f = x(1)^2 + 10 * x(2)^2 - 3 * x(1) * x(2);
end
----------------------------------------------------------------------------------------------------------
and here is the constraints function if that makes any sense in this problem:
function [c, Ceq] = const_hw2q1(x)
c(1) = 2*x(1)+x(2) - 4;
c(2) = x(1) + x(2) + 5;
Ceq = [];
Please help. I know it is a very simple problem but I ran out of options.
  댓글 수: 2
Matt J
Matt J 2020년 12월 2일
편집: Matt J 2020년 12월 2일
It doesn't really make sense for you to be using const_hw2q1() to implement your constraints, when they are in fact linear and could be implemented with appropriate A,b matrices. Also, since your objective is quadratic, quadprog would be a better tool here than fmincon.
Onur Gurler
Onur Gurler 2020년 12월 2일
HI Matt,
I wasn't aware of Quadprog but I'll definitely look into it. Thank you for the suggestion.

댓글을 달려면 로그인하십시오.

채택된 답변

Stephan
Stephan 2020년 12월 2일
편집: Stephan 2020년 12월 2일
Try:
x0 = [0 0];
LB = [-5 -5];
UB = [7 7];
ii = 1;
for jj = 0:0.01:1
xopt(ii,:) = fmincon(@funhw2q1, x0, [], [], [], [], LB, UB, @const_hw2q1);
solution(ii) = xopt(ii,1).^2 + 10 * xopt(ii,2) - 3 * xopt(ii,1) .* xopt(ii,2);
ii = ii + 1;
end
function f = funhw2q1(x)
f = x(1).^2 + 10 * x(2).^2 - 3 * x(1) .* x(2);
end
function [c, Ceq] = const_hw2q1(x)
c(1) = 2*x(1)+x(2) - 4;
c(2) = x(1) + x(2) + 5;
Ceq = [];
end
Note that in every run of your loop you get the same result - what do want to achieve? Since you do not really change anything inside your objective and your constraints, the for loop is not needed and the code simplifies to:
x0 = [0 0];
LB = [-5 -5];
UB = [7 7];
[xopt, solution] = fmincon(@funhw2q1, x0, [], [], [], [], LB, UB, @const_hw2q1)
function f = funhw2q1(x)
f = x(1).^2 + 10 * x(2).^2 - 3 * x(1) .* x(2);
end
function [c, Ceq] = const_hw2q1(x)
c(1) = 2*x(1)+x(2) - 4;
c(2) = x(1) + x(2) + 5;
Ceq = [];
end
  댓글 수: 3
Stephan
Stephan 2020년 12월 2일
Did you notice that you can accept and/or vote for useful answers?
Onur Gurler
Onur Gurler 2020년 12월 2일
I just noticed and gave you a thumbs up.

댓글을 달려면 로그인하십시오.

추가 답변 (0개)

카테고리

Help CenterFile Exchange에서 Surrogate Optimization에 대해 자세히 알아보기

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by