A Stabilized SQP Method : Global Convergence 2 1

Stabilized sequential quadratic programming (SQP) methods for nonlinear optimization are designed to provide a sequence of iterates with fast local convergence regardless of whether or not the active-constraint gradients are linearly dependent. This paper concerns the global convergence properties of a stabilized SQP method with a primal-dual augmented Lagrangian merit function. The proposed method incorporates several novel features. (i) A flexible line search is used based on a direction formed from an approximate solution of a strictly convex QP subproblem and, when one exists, a direction of negative curvature for the primaldual merit function. (ii) When certain conditions hold, an approximate QP solution is computed by solving a single linear system defined in terms of an estimate of the optimal active set. The conditions exploit the formal equivalence between the conventional stabilized SQP subproblem and a bound-constrained QP associated with minimizing a quadratic model of the merit function. (iii) It is shown that with an appropriate choice of termination condition, the method terminates in a finite number of iterations without the assumption of a constraint qualification. The method may be interpreted as an SQP method with an augmented Lagrangian safeguarding strategy. This safeguarding becomes relevant only when the iterates are converging to an infeasible stationary point of the norm of the constraint violations. Otherwise, the method terminates with a point that approximately satisfies certain second-order necessary conditions for optimality. In this situation, if all termination conditions are removed, then the limit points either satisfy the same second-order necessary conditions exactly or fail to satisfy a weak second-order constraint qualification. (iv) The global convergence analysis concerns a specific algorithm that estimates the least curvature of the merit function at each step. If negative curvature directions are omitted, the analysis still applies and establishes convergence to either firstorder solutions or infeasible stationary points. The superlinear convergence of the iterates and the formal local equivalence to stabilized SQP is established in a companion paper (Report CCoM 14-01, Center for Computational Mathematics, University of California, San Diego, 2014).

[1]  Daniel P. Robinson,et al.  A stabilized SQP method: superlinear convergence , 2017, Math. Program..

[2]  Alexey F. Izmailov,et al.  Combining stabilized SQP with the augmented Lagrangian algorithm , 2015, Comput. Optim. Appl..

[3]  A. F. Izmailov,et al.  GLOBALIZING STABILIZED SQP BY SMOOTH PRIMAL-DUAL EXACT PENALTY FUNCTION , 2014 .

[4]  Daniel P. Robinson,et al.  A Globally Convergent Stabilized SQP Method , 2013, SIAM J. Optim..

[5]  Vyacheslav Kungurtsev,et al.  Second-Derivative Sequential Quadratic Programming Methods for Nonlinear Optimization , 2013 .

[6]  Paulo J. S. Silva,et al.  Two New Weak Constraint Qualifications and Applications , 2012, SIAM J. Optim..

[7]  P. Gill,et al.  Sequential Quadratic Programming Methods , 2012 .

[8]  Alexey F. Izmailov,et al.  On attraction of linearly constrained Lagrangian methods and of stabilized and quasi-Newton SQP methods to critical multipliers , 2011, Math. Program..

[9]  Philip E. Gill,et al.  Dynamical Parameter and State Estimation in Neuron Models , 2011 .

[10]  Mikhail V. Solodov,et al.  Stabilized sequential quadratic programming for optimization and a stabilized Newton-type method for variational problems , 2010, Math. Program..

[11]  Daniel P. Robinson,et al.  A primal-dual augmented Lagrangian , 2010, Computational Optimization and Applications.

[12]  José Mario Martínez,et al.  A New Sequential Optimality Condition for Constrained Optimization and Algorithmic Consequences , 2010, SIAM J. Optim..

[13]  José Mario Martínez,et al.  Second-order negative-curvature methods for box-constrained and general constrained optimization , 2010, Comput. Optim. Appl..

[14]  Frank E. Curtis,et al.  Flexible penalty functions for nonlinear constrained optimization , 2008 .

[15]  P. Gill,et al.  State and parameter estimation in nonlinear systems as an optimal tracking problem , 2008 .

[16]  José Mario Martínez,et al.  Augmented Lagrangian methods under the constant positive linear dependence constraint qualification , 2007, Math. Program..

[17]  J. M. Martínez,et al.  On second-order optimality conditions for nonlinear programming , 2007 .

[18]  Daniel P. Robinson Primal -dual methods for nonlinear optimization , 2007 .

[19]  Stephen J. Wright,et al.  Active Set Identification in Nonlinear Programming , 2006, SIAM J. Optim..

[20]  Stephen J. Wright An Algorithm for Degenerate Nonlinear Programming with Rapid Local Convergence , 2005, SIAM J. Optim..

[21]  Javier M. Moguerza,et al.  An augmented Lagrangian interior-point method using directions of negative curvature , 2003, Math. Program..

[22]  Stephen J. Wright Modifying SQP for Degenerate Problems , 2002, SIAM J. Optim..

[23]  Zengxin Wei,et al.  On the Constant Positive Linear Dependence Condition and Its Application to SQP Methods , 1999, SIAM J. Optim..

[24]  J. Gondzio,et al.  Regularized Symmetric Indefinite Systems in Interior Point Methods for Linear and Quadratic Optimization , 1999 .

[25]  William W. Hager,et al.  Stabilized Sequential Quadratic Programming , 1999, Comput. Optim. Appl..

[26]  Stephen J. Wright Superlinear Convergence of a Stabilized SQP Method to a Degenerate Solution , 1998, Comput. Optim. Appl..

[27]  Anders Forsgren,et al.  Primal-Dual Interior Methods for Nonconvex Nonlinear Programming , 1998, SIAM J. Optim..

[28]  Francisco Facchinei,et al.  Convergence to Second Order Stationary Points in Inequality Constrained Optimization , 1998, Math. Oper. Res..

[29]  M. Saunders,et al.  SOLVING REGULARIZED LINEAR PROGRAMS USING BARRIER METHODS AND KKT SYSTEMS , 1996 .

[30]  Paul T. Boggs,et al.  Sequential Quadratic Programming , 1995, Acta Numerica.

[31]  P. Gill,et al.  On the identification of local minimizers in inertia-controlling methods for quadratic programming , 1991 .

[32]  C. B. Luis Une caractérisation complète des minima locaux en programmation quadratique , 1980 .

[33]  Antal Majthay Optimality conditions for quadratic programming , 1971, Math. Program..

[34]  Richard W. Cottle,et al.  On classes of copositive matrices , 1970 .

[35]  Olvi L. Mangasarian,et al.  Nonlinear Programming , 1969 .

[36]  O. Mangasarian,et al.  The Fritz John Necessary Optimality Conditions in the Presence of Equality and Inequality Constraints , 1967 .