Abstract
This paper presents a canonical dual approach for solving a nonconvex global optimization problem governed by a sum of 4th-order polynomial and a log-sum-exp function. Such a problem arises extensively in engineering and sciences. Based on the canonical duality–triality theory, this nonconvex problem is transformed to an equivalent dual problem, which can be solved easily under certain conditions. We proved that both global minimizer and the biggest local extrema of the primal problem can be obtained analytically from the canonical dual solutions. As two special cases, a quartic polynomial minimization and a minimax problem are discussed. Existence conditions are derived, which can be used to classify easy and relative hard instances. Applications are illustrated by several nonconvex and nonsmooth examples.
Similar content being viewed by others
Notes
We use the same definition of the neighborhood as defined in [15] (Note 1 on page 306), i.e., a subset \(\mathcal{X}_0\) is said to be the neighborhood of the critical point \(\bar{\varvec{x}}\) if \(\bar{\varvec{x}}\) is the only critical point in \(\mathcal{X}_0\).
References
Abdi, H., Nahavandi, S.: Designing optimal fault tolerant jacobian for robotic manipulators. In: IEEE Conf. AIM, pp. 426–431 (2010)
Abdi, H., Nahavandi, S., Maciejewski, A.: Optimal fault-tolerant Jacobian matrix generators for redundant manipulators. In: IEEE Int. Conf. Robot., pp. 4688–4693 (2011)
Banichuk, N.: Minimax approach to structural optimization problems. J. Optim. Theory Appl. 20, 111–127 (1976)
Boyd, S., Kim, S., Vandenberghe, L., Hassibi, A.: A tutorial on geometric programming. Optim. Eng. 8, 67–127 (2007)
Boyd, S., Vandenberghe, L.: Convex Optimization. Cambridge University Press, Cambridge (2004)
Chen, Y., Gao, D.: Global solutions of quadratic problems with sphere constraint via canonical dual approach. arXiv:1308.4450 (2013)
Chiang, M.: Geometric Programming for Communication Systems. Now Publishers Inc, Hanover (2005)
Chiang, M., Boyd, S.: Geometric programming duals of channel capacity and rate distortion. IEEE Trans. Inf. Theory 50, 245–258 (2004)
Chiang, M., Tan, C., Palomar, D., O’Neill, D., Julian, D.: Power control by geometric programming. IEEE Trans. Wirel. Commun. 6, 2640–2651 (2007)
Fang, S., Gao, D., Sheu, R., Wu, S.: Canonical dual approach to solving 0–1 quadratic programming problems. J. Ind. Manag. Optim. 4, 125–142 (2008)
Gao, D.: Minimax and triality theory in nonsmooth variational problems. In: Fukushima, M., Liqun, Qi. (eds.) Reformulation-Nonsmooth, Piecewise Smooth, Semismooth and Smoothing Methods, pp. 161–180 (1998)
Gao, D.: Canonical dual transformation method and generalized triality theory in nonsmooth global optimization. J. Global Optim. 17, 127–160 (2000)
Gao, D.: Duality principles in nonconvex systems: theory, methods, and applications. Springer, Netherlands (2000)
Gao, D.: Finite deformation beam models and triality theory in dynamical post-buckling analysis. Int. J. Nonlinear Mech. 35, 103–131 (2000)
Gao, D.: Nonconvex semi-linear problems and canonical duality solutions. Adv. Mech. Math. 2, 261–312 (2003)
Gao, D.: Canonical duality theory and solutions to constrained nonconvex quadratic programming. J. Global Optim. 29, 377–399 (2004)
Gao, D.: Complete solutions and extremality criteria to polynomial optimization problems. J. Global Optim. 35, 131–143 (2006)
Gao, D., Ruan, N.: Solutions to quadratic minimization problems with box and integer constraints. J. Global Optim. 47, 463–484 (2010)
Gao, D., Ruan, N., Pardalos, P.: Canonical dual solutions to sum of fourth-order polynomials minimization problems with applications to sensor network localization. In: Boginski, V. L., Commander, C. W., Pardalos, P. M., Ye, Y. (eds.) Sensors: Theory, Algorithms, and Applications. Springer, Berlin, pp. 37–54 (2012)
Gao, D., Ruan, N., Sherali, H.: Solutions and optimality criteria for nonconvex constrained global optimization problems with connections between canonical and Lagrangian duality. J. Global Optim. 45, 473–497 (2009)
Gao, D., Ruan, N., Sherali, H.: Canonical dual solutions for fixed cost quadratic programs. In: Chinchuluun, A., Pardalos, P. M., Enkhbat, R., Tseveendorj, I. (eds.) Optimization and Optimal Control. Springer, Berlin, pp. 139–156 (2010)
Gao, D., Watson, L., Easterling, D., Thacker, W., Billups, S.: Solving the canonical dual of box- and integer-constrained nonconvex quadratic programs via a deterministic direct search algorithm. Optim. Methods Softw. 26(1), 1–14 (2011)
Gao, D., Wu, C.: On the triality theory for a quartic polynomial optimization problem. J. Ind. Manag. Optim. 8, 229–242 (2012)
Kiwiel, K.: Methods of Descent for Nondifferentiable Optimization. Springer, Berlin (1985)
Pee, E., Royset, J.: On solving large-scale finite minimax problems using exponential smoothing. J. Optim. Theory Appl. 148, 390–421 (2011)
Polak, E.: Optimization: Algorithms and Consistent Approximations. Springer, Berlin (1997)
Polak, E., Royset, J., Womersley, R.: Algorithms with adaptive smoothing for finite minimax problems. J. Optim. Theory App. 119, 459–484 (2003)
Roberts, R., Yu, H.G. Maciejewski, A.: Characterizing optimally fault-tolerant manipulators based on relative manipulability indices. In IEEE Conf. IROS, pp. 3925–3930 (2007)
Royset, J., Polak, E., Kiureghian, A.: Adaptive approximations and exact penalization for the solution of generalized semi-infinite min-max problems. SIAM J. Optim. 14, 1–34 (2004)
Strang, G.: A minimax problem in plasticity theory. In: Nashed, M. Z. (ed.) Functional Analysis Methods in Numerical Analysis. Springer, Berlin, pp. 319–333 (1979)
Wang, Z., Fang, S., Gao, D., Xing, W.: Canonical dual approach to solving the maximum cut problem. J. Global Optim. 54(2), 341-351 (2012)
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
The following lemma is a generalization of Lemma 6 in [23].
Lemma 7
Suppose that \( \varvec{P}\in \mathbb {R}^{n\times n}\), \( \varvec{U}\in \mathbb {R}^{m\times m}\) and \( \varvec{D}\in \mathbb {R}^{n\times m}\) are given symmetric matrices with
where \( \varvec{P}_{11}\), \( \varvec{U}_{11}\) and \( \varvec{D}_{11}\) are \(r\times r\)-dimensional matrices, and \( \varvec{D}_{11}\) is nonsingular. Then,
Proof
Obviously, \( \varvec{P}+ \varvec{D}\varvec{U}\varvec{D}^T\preceq 0\) is equivalent to
By Schur lemma, Eq. (47) is equivalent to
The inverse of matrix \( \varvec{P}\) is
Then, it is easy to prove that \(- \varvec{P}_{11}+ \varvec{P}_{12} \varvec{P}_{22}^{-1} \varvec{P}_{21}\succ 0\). Since \( \varvec{D}_{11}\) is nonsingular and \( \varvec{U}_{11}\succ 0\), we have \( \varvec{D}_{11} \varvec{U}_{11} \varvec{D}_{11}^T\succ 0\). Thus the Eq. (48) is equivalent to
which is further equivalent to
Since \( \varvec{D}_{11}^T(- \varvec{P}_{11}+ \varvec{P}_{12} \varvec{P}_{22}^{-1} \varvec{P}_{21})^{-1} \varvec{D}_{11}=- \varvec{D}^T \varvec{P}^{-1} \varvec{D}\) and \( \varvec{U}_{22}\succ 0\), the Eq. (50) is equivalent to
The lemma is proved. \(\square \)
Rights and permissions
About this article
Cite this article
Chen, Y., Gao, D.Y. Global solutions to nonconvex optimization of 4th-order polynomial and log-sum-exp functions. J Glob Optim 64, 417–431 (2016). https://doi.org/10.1007/s10898-014-0244-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10898-014-0244-5
Keywords
- Global optimization
- Canonical duality theory
- Double-well function
- Log-sum-exp function
- Polynomial minimisation
- Minimax problems