Abstract
We will show examples in which the primal sequence generated by the Newton–Lagrange method converges to a strict local minimizer of a constrained optimization problem but the gradient of the Lagrangian does not tend to zero, independently of the choice of the dual sequence.
References
Andreani, R., Haeser, G., Martínez, J.M.: On sequential optimality conditions for smooth constrained optimization. Optimization 60, 627–641 (2011)
Andreani, R., Martínez, J.M., Santos, L.T., Svaiter, B.F.: On the behavior of constrained optimization methods when Lagrange multipliers do not exist. Optim. Methods Softw. 29, 646–657 (2014)
Andreani, R., Martínez, J.M., Svaiter, B.F.: A new sequential optimality condition for constrained optimization and algorithmic consequences. SIAM J. Optim. 20, 3533–3554 (2010)
Birgin, E.G., Martínez, J.M.: Practical Augmented Lagrangian Methods for Constrained Optimization. SIAM, Philadelphia (2014)
Fernández, D., Solodov, M.V.: Stabilized sequential quadratic programming: a survey. Pesquisa Oper. 34, 591–604 (2013)
Izmailov, A., Solodov, M.V.: Stabilized SQP revisited. Math. Program. 133, 93–120 (2012)
Izmailov, A.F., Solodov, M.V., Uskov, E.I.: Combining stabilized SQP with the augmented Lagrangian algorithm. Comput Optim Appl. 62, 405–429 (2015)
Izmailov, A.F., Solodov, M.V., Uskov, E.I.: Globalizing stabilized sequential quadratic programming method by smooth primal-dual exact penalty function. J Optim Theor Appl, 1–31 (2016)
Martínez, J.M., Svaiter, B.F.: A practical optimality condition without constraint qualifications for nonlinear programming. J. Optim. Theory Appl. 118, 117–133 (2003)
Nocedal, J., Wright, S.J.: Numerical Optimization, 2nd edn. Springer, Berlin (2006)
Wright, S.J.: Superlinear convergence of a stabilized SQP method to a degenerate solution. Comput. Optim. Appl. 11, 253–275 (1998)
Wright, S.J.: Modifying SQP for degenerate problems. SIAM J. Optim. 13, 470–497 (2002)
Acknowledgments
We wish to acknowledge the associate editor and two anonymous referees for insightful comments that helped us a lot to improve this paper. This work was kindly supported by the Brazilian research agencies CNPq (303013/2013–3, 305078/2013–5), PRONEX-CNPq/FAPERJ (E–26/111.449/2010–APQ1), and FAPESP (2010/10133–0, 2013/05475–7, 2013/07375–0).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Andreani, R., Martínez, J.M. & Santos, L.T. Newton’s method may fail to recognize proximity to optimal points in constrained optimization. Math. Program. 160, 547–555 (2016). https://doi.org/10.1007/s10107-016-0994-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10107-016-0994-6