| Title |
Structural Optimization Algorithm based on Neural Dynamics Model |
| Abstract |
A structural optimization algorithm based on neural dynamics model is presented in this paper. Apseudo-objective function for generally constrained optimization problems is formunlated by the exterior penalty function method. The neural dynamics model is developed in the form of a system of differential equations from the Lyapunov stability theorem. The Kuhn-Tucker necessary conditions for a local minimum are adopted to check the equilibrium point obtained from the nerual dynamics model. The topology of the neural dynamics model consists of two primary layer: variable layer and constraint layer. The numbers of modes in the variable constraint layers correspond to the numbers of design variables and constraints in the optimization problem. The global convergence and robustness of the model are guaranteed by integrating the stability theorem, Kuhn-Tucker conditions, and exterior penalty function method. In a companion paper the neural dynamics model is applied to optimum design of steel structure subject to both displacment and stress constraints. |