Discrete-Time ZNN Algorithms for Time-Varying Quadratic Programming Subject to Time-Varying Equality Constraint

Abstract

A special class of recurrent neural network (RNN), i.e., Zhang neural network (ZNN), has been proposed for a decade for solving online various time-varying problems. In this paper, we generalize and investigate a continuous-time ZNN model for online solution of the time-varying convex quadratic programming (QP) subject to a time-varying linear equality constraint. For the purpose of possible hardware (e.g., digital-circuit or digital-computer) realization, discrete-time ZNN models and numerical algorithms (i.e., discrete-time ZNN algorithms, in short) are proposed and developed by using Euler difference rules. Computer-simulation and numerical results demonstrate the efficacy and accuracy of the presented continuous-time ZNN model and the proposed discrete-time ZNN algorithms for solving online time-varying QP problems.