Solving convex quadratic programming problems by an modified neural network with exponential convergence

Abstract

This paper presents using a modified neural network with exponential convergence to solve strictly quadratic programming problems with general linear constraints. It is shown that the proposed neural network is globally convergent to a unique optimal solution within a finite time. Compared with the existing the primal-dual neural network and the dual neural network for solving such problems, the proposed neural network has a low complexity for implementation and can be guaranteed to have a exponential convergence rate.