Some necessary and some sufficient conditions for local extrema of polynomials and power series in two variables

V. N. Nefedov

Journal of Samara State Technical University, Ser. Physical and Mathematical Sciences, Tome 28 (2024) no. 4, pp. 615-650.

Résumé

This study extends the author's previous works establishing necessary and sufficient conditions for a local extremum at a stationary point of a polynomial or an absolutely convergent power series in its neighborhood. It is known that in the one-dimensional case, the necessary and sufficient conditions for an extremum coincide, forming a single criterion. The next stage of analysis focuses on the two-dimensional case, which constitutes the subject of the present research. Verification of extremum conditions in this case reduces to algorithmically feasible procedures: computing real roots of univariate polynomials and solving a series of practically implementable auxiliary problems. An algorithm based on these procedures is proposed. For situations where its applicability is limited, a method of substituting polynomials with undetermined coefficients is developed. Building on this method, an algorithm is constructed to unambiguously verify the presence of a local minimum at a stationary point for polynomials representable as a sum of two $A$-quasihomogeneous forms, where $A$ is a two-dimensional vector with natural components.

Détail

Keywords: polynomials, power series, necessary conditions for an extremum, sufficient conditions for an extremum, quasi-homogeneous forms

Bibliographie

[1] Nefedov V. N., “Estimation of the error in convex polynomial optimization problems”, U.S.S.R. Comput. Math. Math. Phys., 30:1 (1990), 147–158 | DOI | MR | Zbl

[2] Nefedov V. N., “Necessary and sufficient conditions for an extremum in complex problems of optimization of systems described by polynomial and analytic functions”, J. Comput. Syst. Sci. Int., 62:2 (2023), 179–200 | DOI | DOI

[3] Vasil'ev F. P., Metody optimizatsii [Optimization Methods], v. 1, Konechnomernye zadachi optimizatsii. Printsip maksimuma. Dinamicheskoe programmirovanie [Finite-dimensional Optimization Problems. Maximum Principle. Dynamic Programming], MTsNMO, Moscow, 2011, 620 pp. (In Russian)

[4] Krasnosel'skiy M. A., Vainikko G. M., Zabreiko P. P., et al., Approximate Solution of Operator Equations, Wolters-Noordhoff Series of Monographs and Textbooks on Pure and Applied Mathematics, Wolters-Noordhoff Publ., Groningen, 1972, xii+484 pp. | Zbl

[5] Nefedov V. N., “Necessary and sufficient conditions of the extremum in the analytical optimization problems”, Trudy MAI, 2009, no. 33, 4 (In Russian)

[6] Gindikin S. G., “Energy estimates connected with the Newton polyhedron”, Tr. Mosk. Mat. Obs., 31 (1974), 189–236 (In Russian) | MR | Zbl

[7] Bruno A. D., Power Geometry in Algebraic and Differential Equations, North-Holland Mathematical Library, 57, North-Holland, Amsterdam, 2000, ix+385 pp. | Zbl | Zbl

[8] Gindikin S. G., Volevich L. R., The Method of Newton's Polyhedron in the Theory of Partial Differential Equations, Mathematics and Its Applications. Soviet Series, 86, Kluwer Academic Publ, Dordrecht, 1992, x+266 pp. | Zbl

[9] Khovansky A. G., “Polyhedra and algebra”, Trudy ISA RAN, 38 (2008), 23–35 (In Russian)

[10] Nefedov V., “Methods and algorithms for determining the main quasi-homogeneous forms of polynomials and power series”, MATEC Web of Conferences, 362 (2022), 01017 | DOI

[11] Nefedov V. N., “On one method of analysis of a polynomial on constancy of signs in the positive orthant”, Trudy MAI, 2006, no. 22, 6 (In Russian)

[12] Evtushenko Yu. G., Rat'kin V. A., “The method of half-divisions for global optimization of a function of many variables”, Sov. J. Comput. Syst. Sci., 25:5 (1987), 75–84 | Zbl