A new nonlinear neural networks based on exact penalty function with two-parameter

The penalty function method is one of the most important ways to solve mathematical programming. Its main idea is to transform a constrained mathematical programming into a sequential unconstrained mathematical programming which is easier to solve. In the paper, first, we introduce an exact penalty function with two-parameter. Based on the penalty, it makes the construction of energy function easy, so we can import a new nonlinear neural network based on this kinds of energy function. Secondly we probe into the stability of such system and the relationship between the equilibrium points and the energy function. It is shown that under some proper given conditions, an optimal solution of the nonlinear programming problems is an equilibrium point of the neural dynamics. Finally we give two examples.