A dynamic system model for solving a class of geometric programming problems

In this article, we investigate how to use neural networks to solve geometric programming (GP) problems efficiently. The main idea is to convert the GP problem into an equivalent convex optimization problem. A neural network model is then constructed for solving the obtained convex programming problem. Its Lyapunov stability and global convergence are proved under some conditions. The simulation results also show that the proposed neural network is feasible and efficient.