<div dir="ltr">Hi everyone,<div><br></div><div>I have been having a great deal of success with Ipopt, solving large scale non-linear problems in Optimal Control and I wanted to thank everyone involved for doing such a great job with this package-we will of course site it in our upcoming papers.</div><div><br></div><div>I had a small issue recently with the restoration phase. The problems are very similar with the communication back in 2010, here</div><div><b><a href="http://list.coin-or.org/pipermail/ipopt/2010-January/001855.html">http://list.coin-or.org/pipermail/ipopt/2010-January/001855.html</a></b></div><div><b><br></b></div><div>The error is </div><div><b>Restoration phase is called at point that is almost feasible,</b></div><div><b> with constraint violation 4.590405e-05. Abort.</b></div><div><b><br></b></div><div>Derivative checker shows no errors with derivatives. I use Hessian approximation </div><div>app->Options()->SetStringValue("hessian_approximation", "limited-memory");<br></div><div><br></div><div>Cross-checking with another solver the final solution appears correct, within my desired tolerance.</div><div><b><br></b></div><div>I haven't been able to find whether there is a solution to this "issue" other than perhaps providing an exact Hessian calculation instead of using BFGS.<br clear="all"><div><br></div><div>If anyone has any new information about this issue, could you please share it?</div><div><br></div><div>Warm regards, </div>-- <br><div class="gmail_signature"><div dir="ltr"><div><div dir="ltr">Panos Lambrianides<div><a href="mailto:panos@soe.ucsc.edu" target="_blank">panos@soe.ucsc.edu</a></div><div>(415) 713-6718 (Mobile)<br></div><div><a href="https://sites.google.com/a/ucsc.edu/uncertainty-quantification/" target="_blank">https://sites.google.com/a/ucsc.edu/uncertainty-quantification/</a><br></div><div><br></div><div><br></div></div></div></div></div>
</div></div>