Feedback stabilization methods for the solution of nonlinear programming problems (Q398644)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Feedback stabilization methods for the solution of nonlinear programming problems |
scientific article |
Statements
Feedback stabilization methods for the solution of nonlinear programming problems (English)
0 references
15 August 2014
0 references
\noindent The goal of the paper is to show the construction of a well-defined dynamical system on the feasible set of a nonlinear programming problem (NLP) with the following properties:\newline P 1: The conditions of Nagumo's theorem, given on p. 27 of \textit{J.-P. Aubin} [Viability theory. Boston, MA etc.: Birkhäuser (1991; Zbl 0755.93003)] must be satisfied.\newline P 2: The vector field appearing in the right-hand side of the dynamical system is a locally Lipschitz vector field.\newline P 3: The equilibrium points of the dynamical system are exactly the points for which the necessary Karush-Kuhn-Tucker conditions of the NLP hold.\newline P 4: The objective function of the NLP is a (strict) Lyapunov function for the dynamical system.\newline P 5: The vector field appearing in the right-hand side of the dynamical system must be explicitly known.\newline P 6: The vector field appearing in the right-hand side of the dynamical system must have free parameters, which can be selected in an appropriate way so that the convergence properties of the corresponding numerical schemes to the global attractor of the dynamical system are optimal.\newline The author claims that the theory used for the construction of the algorithm is different from other existing algorithms, although other algorithms have much better characteristics than the algorithms proposed in this paper. So, no claim is made about the effectiveness of the proposed algorithms. Moreover, no claim is made about the generality of the results: the linear independence constraint qualification assumed in this work is a restrictive assumption. The paper provides some examples, which show the performance of the algorithms.
0 references
nonlinear programming
0 references
feedback stabilization
0 references
Lyapunov functions
0 references
nonlinear systems
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references