Using limitation to get two ranges of a state

[劉宇軒]

I have a problem with the casadi and ipopt.

When I try to limit the range of a state, I try to use the limitation. For example, the value range of a variable x is [0,1], I want to change the value range to [0,0.4] and [0.6,1] by limiting (x-0.5)^2>0.1^2, but the range from the solver is only available in [0,0.4];

Can I get the range I wanted?

and the Matlab code is:

import casadi.*

x=MX.sym('x',1);

x0=0;

lbx=[0];

ubx=[1];

g=[(x-0.5)^2];

lbg=[0.1^2]

ubg=1;

J=1e11*(x-0.9)^2;

nlp = struct('x', x, 'f', J, 'g', g);

options = struct;

options.ipopt.print_level = false;

options.print_time=false;

options.ipopt.max_iter = 15000;

solver = casadi.nlpsol('solver', 'ipopt', nlp,options);

sol = solver('x0', x0, 'lbx', lbx, 'ubx', ubx,'lbg', lbg, 'ubg', ubg);

u_opt = full(sol.x)