How to Solve Semi Definite Linear Matrix Inequalities in MATLAB Builtin LMI Toolbox? How trace of a matrix or log(determinant(matrix)) can be taken as objective in MATLAB Builtin LMI toolbox?
I am not sure about built in toolbox. I would recommend Yalmip toolbox together with MATLAB. You can find it in the link https://yalmip.github.io/
Log det objectives are possible.
Helo Owais Khan
I used yalmip and SeDuMi. This public domain and I recomended this, I used this in Matlab 2016b. Could I help you??
Have a nice day
Yalmip with SeDuMi is a great tool but sometimes I find the slower MATLAB tool to be more stable. It handles both logdet and trace objectives in a seamless way, here you can find a working example https://it.mathworks.com/help/robust/ug/advanced-topics.html
Monika Rybczak Thank you sou much. sorry for my late reply. i was actually busy. so nice of you to offer me your kind help. i have already worked on Yalmip . Actually i am trying to reproduce one of Robust MPC paper using LMIs in which the author produce the results using Built-in LMI toolbox. Could you please help me to plot such 3D plots attached here in which state x is unmeasurable and error e is uncertain but its bound is refreshed at each instant.
An update. In recent days I've compared lots of solvers for this kind of problems. The one which overcome others was Yalmip + SDPT3. Another solution is approximating lodget with trace, which introduces a significant approximation. A better approximation is done by SeDuMi, using a longer series expansion. Other Yalmip solvers use the geomean (geometric mean of eigenvalues), which works quite well, but still SDPT3 seems to be better at solving -logdet (volume maximization) problems under LMI constraints.
- Badges
- Science topic
- Similar topics
- Mathematics
- Algebra
- Matrix