Abstract
The augmented Lagrangian SQP subroutine OPALQP was originally designed for small-to-medium sized constrained optimization problems in which the main calculation on each iteration, the solution of a quadratic program, involves dense, rather than sparse, matrices. In this paper, we consider some reformulations of OPALQP which are better able to take advantage of sparsity in the objective function and constraints.
The modified versions of OPALQP differ from the original in using sparse data structures for the Jacobian matrix of constraints and in replacing the dense quasi-Newton estimate of the inverse Hessian of the Lagrangian by a sparse approximation to the Hessian. We consider a very simple sparse update for estimating ∇2 L and also investigate the benefits of using exact second derivatives, noting in the latter case that safeguards are needed to ensure that a suitable search direction is obtained when ∇2 L is not positive definite on the null space of the active constraints.
Similar content being viewed by others
References
Murtagh, B. A., andSaunders, M. A.,A Projected Lagrangian Algorithm and Its Implementation for Sparse Nonlinear Constraints, Mathematical Programming Study, Vol. 16, pp. 84–117, 1982.
Smith, S., andLasdon, L.,Solving Large Sparse Nonlinear Programs Using GRG, ORSA Journal on Computing, Vol. 4, pp. 1–15, 1992.
Conn, A., Gould, N. I. M., andToint, P.,A Comprehensive Description of LANCELOT, Technical Report 91/10, Facultés Universitaires de Namur, 1991.
Anonymous, N. N.,OPTIMA User's Guide, Numerical Optimisation Centre, University of Hertfordshire, 1988.
Bartholomew-Biggs, M. C.,The Development of Recursive Quadratic Programming Methods Based on the Augmented Lagrangian, Mathematical Programming Study, Vol. 31, pp. 21–41, 1987.
Bartholomew-Biggs, M. C., andNguyen, T. T.,A Local Convergence Analysis for REQP Using Conjugate Basis Matrices, Journal of Optimization Theory and Applications, Vol. 71, pp. 31–45, 1990.
Conn, A. R., Gould, N. I. M., andToint, P.,Large-Scale Nonlinear Constrained Optimization, Proceedings of the Second International Conference on Industrial and Applied Mathematics, Edited by R. E. O'Malley, Jr, SIAM, Philadelphia, Pennsylvania, pp 51–70, 1992.
Bartholomew-Biggs, M. C., andHernandez, M.,Some Improvements to Subroutine OPALQP for Dealing with Large Problems, Journal of Economic Dynamics and Control, Vol. 18, pp. 185–203, 1994.
Toint, P.,On Sparse and Symmetric Updating Subject to a Linear Equation, Mathematics of Computation, Vol. 31, pp. 954–961, 1977.
Toint, P.,A Note on Sparsity-Exploiting Quasi-Newton Methods, Mathematical Programming, Vol. 21, pp. 172–181, 1981.
Fletcher, R.,An Optimal Positive-Definite Update for Sparse Hessian Matrices, Report NA/145, Department of Mathematics, University of Dundee, 1993.
Mahidhara, D., andLasdon, L.,An SQP Algorithm for Large, Sparse, Nonlinear Problems, School of Business Administration, University of Texas, 1990.
Duff, I. M., andReid, J. K.,MA27: A Set of Fortran Subroutines for Solving Sparse Symmetric Sets of Linear Equations, Report R-10533, Atomic Energy Research Establishment, Harwell, England, 1982.
Duff, I. M., Reid, J. K. Scott, J. A., andTurner, K.,The Factorization of Sparse Symmetric Indefinite Matrices, Report RAL-90-066, Rutherford Appleton Laboratory, 1990.
Schnabel, R., andEskow, E.,A New Modified Cholesky Factorization, Report CU-CS-415-88, University of Colorado, 1988.
Forsgren, A., andMurray, W.,Newton Methods for Large-Scale Linear Equality Constrained Minimization, Report SOL 90-6, Stanford University, 1990.
Gould, N. I. M.,On the Accurate Determination of Search of Search Directions for Simple Differentiable Penalty Functions, IMA Journal of Numerical Analysis, Vol. 6, pp. 357–372, 1986.
Author information
Authors and Affiliations
Additional information
Communicated by L. C. W. Dixon
The authors are grateful to John Reid and Nick Gould of the Rutherford Appleton Laboratory for a number of helpful and interesting discussions. Thanks are also due to Laurence Dixon for comments which led to the clarification of some parts of the paper.
This work has been partly supported by a CAPES Research Studentship funded by the Brazilian Government.
Rights and permissions
About this article
Cite this article
Bartholomew-Biggs, M.C., Hernandez, M.D.F.G. Using the KKT matrix in an augmented Lagrangian SQP method for sparse constrained optimization. J Optim Theory Appl 85, 201–220 (1995). https://doi.org/10.1007/BF02192305
Issue Date:
DOI: https://doi.org/10.1007/BF02192305