Summary
Recently developed projected Newton methods for minimization problems in polyhedrons and Cartesian products of Euclidean balls are extended here to general convex feasible sets defined by finitely many smooth nonlinear inequalities. Iterate sequences generated by this scheme are shown to be locally superlinearly convergent to nonsingular extremals\(\bar u\), and more specifically, to local minimizers\(\bar u\) satisfying the standard second order Kuhn-Tucker sufficient conditions; moreover, all such convergent iterate sequences eventually enter and remain within the smooth manifold defined by the active constraints at\(\bar u\). Implementation issues are considered for large scale specially structured nonlinear programs, and in particular, for multistage discrete-time optimal control problems; in the latter case, overall per iteration computational costs will typically increase only linearly with the number of stages. Sample calculations are presented for nonlinear programs in a right circular cylinder in ℝ3.
Similar content being viewed by others
References
Bertsekas, D.: Projected Newton methods for optimization problems with simple constraints. SIAM J. Control Optimization20, 221–246 (1982)
Gafni, E.M., Bertsekas, D.: Two-metric projection methods for constrained optimization. SIAM J. Control Optimization22, 936–964 (1984)
Gawande, M.: Projected gradient and projected Newton algorithms for specially structured constrained minimization problems. Ph.D. Dissertation, North Carolina State University 1986
Gawande, M., Dunn, J.C.: A projected Newton method in a Cartesian product of balls. J. Optimization Theory Appl.59 (1) (1988) (to appear)
Ortega, J.M., Rheinboldt, W.C.: Iterative solution of nonlinear equations in several variables. New York: Academic Press 1970
Dunn, J.C.: On the convergence of projected gradient processes to singular critical points. J. Optimization Theory Appl.55, (2) 203–216 (1987)
Moré, J.J., Burke, J.V.: On the identification of active constraints. Report No. 82, Mathematics and Computer Science Division. Argonne National Laboratory 1986
Dunn, J.C.: Gradient projection methods for systems optimization problems. In: Advances in Control and Dynamic Systems (C.T. Leondes, ed.), London: Academic Press 1988
Goldstein, A.A.: On Newton's method. Numer. Math.7, 391–393 (1965)
Levitin, E.S., Poljak, B.T.: Constrained optimization methods. USSR Comput. Math. Math. Phys.6, 1–50 (1966)
Bertsekas, D.: On the Goldstein-Levitin-Poljak gradient projection method. IEEE Trans. Autom. Contr., AC21, 174–184 (1976)
Dunn, J.C.: Global and asymptotic convergence rate estimates for a class of projected gradient processes. SIAM J. Control Optimization19, 368–400 (1980)
Dunn, J.C.: Newton's method and the Goldstein step length rule for constrained minimization problems. SIAM J. Control Optimization18, 659–674 (1980)
Hughes, G.C., Dunn, J.C.: Newton-Goldstein convergence rates for convex constrained minimization problems with singular solutions. Appl. Math. Optimization12, 203–230 (1984)
Rustem, B.: A class of superlinearly convergent projection algorithms with relaxed step sizes. Appl. Math. Optimization12, 29–43 (1984)
Sachs, E.: Newton's method for singular constrained optimization problems. Appl. Math. Optimization11, 247–276 (1984)
Gawande, M., Dunn, J.C.: Variable metric gradient projection processes in convex feasible sets defined by nonlinear inequalities. Appl. Math. Optimization17 (2), 103–119 (1988)
Dennis, J.E., Jr., Schnabel, R.B.: Numerical methods for unconstrained optimization and nonlinear equations. Englewood Cliffs, N.J.: Prentice-Hall 1983
Decker, D.W., Keller, H.B., Kelley, C.T.: Convergence rates for Newton's method at singular points. SIAM J. Numer. Anal.20, 296–313 (1983)
Dunn, J.C.: Efficient dynamic programming implementation of Newton's method for unconstrained optimal control problems. J. Optimization Theory Appl. (to appear)
Wilson, R.B.: A simplicial algorithm for concave programming. Ph.D. Dissertation, Harvard University 1963
Tapia, R.A.: A stable approach to Newton's method for general mathematical programming problems inR n. J. Optimization Theory Appl.14, 453–476 (1974)
Garcia-Palomares, U.M., Mangasarian, O.L.: Superlinearly convergent quasi-Newton algorithms for nonlinearly constrained optimization problems. Math. Program11, 1–13 (1976)
Han, S-P.: Superlinearly convergent variable metric algorithms for general nonlinear programming problems. Math. Program11, 263–282 (1976)
Tapia, R.A.: Diagonalized multiplier methods and quasi-Newton methods for constrained optimization. J. Optimization Theory Appl.22, 135–194 (1977)
Murray, W., Wright, M.H.: Projected Lagrangian methods based on the trajectories of penalty and barrier functions. Syst. Optimization Lab. Report 78-23, Stanford University 1978
Gill, P.E., Murray, W.: The computation of Lagrange multiplier estimates for constrained minimization. Math. Program17, 32–60 (1979)
Fletcher, R.: Practical methods of optimization, Vol. 2. New York: Wiley 1981
Boggs, P.T., Tolle, J.W., Wang, P.: On the local convergence of quasi-Newton methods for constrained optimization. SIAM J. Control Optimization20, 161–171 (1982)
Nocedal, J., Overton, M.L.: Projected Hessian updating algorithms for nonlinearly constrained optimization. SIAM J. Numer. Anal.22, 821–850 (1985)
Gill, P.E., Murray, W., Saunders, M.A., Wright, M.H.: Stanford University, Department of Operations Research. Report No. SOL 86-6R, 1986
Byrd, R.H., Schnabel, R.B.: Continuity of the null space basis and constrained optimization. Math. Program35, 32–41 (1986)
Vulikh, B.Z.: Introduction to functional analysis for scientists and technologists. New York: Pergamon Press 1963
Dunn, J.C.: Asymptotic decay rates from the growth properties of Liapunov functions near singular attractors. J. Math. Anal. Appl.125(1), 6–21 (1987)
Polak, E.: Computational methods in optimization. New York: Academic Press 1971
Murray, D.M., Yakowitz, S.: Differential dynamic programming and Newton's method for diserete optimal control problems. J. Optimization Theory Appl.43, 395–414 (1984)
Ohno, K.: Differential dynamic programming and separable programs. J. Optimization Theory Appl.24, 617–637 (1978)
Drud, A.: CONOPT: A GRG Code for large sparse dynamic nonlinear optimization Problems. Math. Program31, 153–191 (1985)
Author information
Authors and Affiliations
Additional information
Investigation supported by NSF Research Grant #DMS-85-03746
Rights and permissions
About this article
Cite this article
Dunn, J.C. A projected Newton method for minimization problems with nonlinear inequality constraints. Numer. Math. 53, 377–409 (1988). https://doi.org/10.1007/BF01396325
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF01396325