New approach for solving optimization problems in economic load dispatch using Hopfield neural networks

Hopfield neural networks are extended to handle inequality constraints by introducing the penalty function method. The use of the penalty function is to convert constrained optimization problems into unconstrained problems. The Hopfield neural network can only be applied for the optimization with linear equality and inequality constraints. Due to the fact that problem involved here involves quadratic equality constraints, they need to convert them into linear constraints using the Taylor's series expansion. This paper presents an alternative method to extend the Hopfield neural networks for solving inequality constraints and converting the quadratic equality constraints into the linear form for the application to the economic load dispatch problem.