Open cnpcshangbo opened 5 years ago
crossover frequency has been met, but the phase margin hasn't.
I already tried to tune up the parameter M1, however, it doesn't work even I put it 10^1000.
Try multi-objective optimization in MATLAB. Two options:
For optimizing a single objective function subject to constraints, see Constrained Optimization.
For optimizing multiple objective functions, see Multiobjective Optimization.
https://ww2.mathworks.cn/help/optim/nonlinear-programming.html
I'll try constrained optimization first. https://ww2.mathworks.cn/help/gads/examples/constrained-minimization-using-pattern-search.html
4127c7905cb145b41521a78799dd0b08263b6fd3 Optimization terminated: no feasible point found. To do: use the values for ACC paper.
% k_p=x(1)=2.1654;
% k_i=x(2)=0;
% k_d=x(3)=2.1654*0.6192;
% lambda=x(4)=0;
% mu=x(5)=1.0497;
The parameters make the Bode plot magnitude curve look good. However, the U_c1 and U_c2 are not zero. Therefore, there should be something wrong with U_c1 and U_c2.
83c0bd342b8bd6b7b40e7353beb5263ff71cf77f
to do:
check the formula for U_c1 and U_c2.
Two errors found:
G(jw) was set to zero previously.
Optimization terminated: mesh size less than options.MeshTolerance try multi-obj