Artificial Neural System Method for Solving Nonlinear Programming with Linear Equality Constraints

A new artificial neural system model for solving nonlinear programming with equality constraints is proposed in this paper. This model has two properties as follows: first, the optima set to the problems coincides with the set of equilibria of the neural system model which means the proposed model is complete, second, the model converges globally to an exact optimal solution of the nonlinear programming for any starting point from the feasible region. Compared with the existing models, these two properties indicate that the proposed model is more competitive and thus a novel neural system method for solving nonlinear programming with equality constraints.