The equations are highly non-linear, and there could be multiple solutions. Is there a way to impose specific constraints on the solution?
Also, in many cases, it is very hard to converge to exact solution, so if MATLAB is used, how to set tolerances for solutions?
Hi
I advice you to see these documents with some examples. You will find what you need.
Best regard
http://www.math.umn.edu/~olver/am_/odz.pdf
http://fr.mathworks.com/help/symbolic/solve-an-algebraic-equation.html
"fsolve" can solve systems of nonlinear equations. You have to provide a starting point as a guess. Algorithms, tolerances, etc., can be set using "optimset". Typing "help fsolve" will reveal more information.
If constraints are to be set, then "fsolve" is not applicable. In that case, "fmincon" may be used. This is for constrained minimization. Once more a starting guess is required. Also, you need to set an objective function to be minimized.
If there are multiple solutions, then these algorithms will typically find the one closest to their starting point. The only sure way I know of to find a global optimum is to use a genetic algorithm.
Hi
FSOLVE solves systems of nonlinear equations of several variables.
Reference: Type 'help fsolve' or 'doc fsolve' in the MATLAB command window.
Download fsolveDemo.m
Best regards
http://www.dam.brown.edu/people/sahn/fsolveDemo.m
- Badges
- Science topic