Global optimization methods for designing and training neural networks

2003 ◽  
Author(s):  
A. Yamazaki ◽  
T.B. Ludermir ◽  
M.C.P. de Souto
Keyword(s):  
Neural Networks ◽  
Global Optimization ◽  
Optimization Methods ◽  
Global Optimization Methods ◽  
And Training

Identification of the Optimal Product Configuration and Parameters for Improving Manufacturability

1996 ◽  
Author(s):  
Deyi Xue
Keyword(s):  
Global Optimization ◽  
Optimization Methods ◽  
Product Configuration ◽  
Design System ◽  
Optimization Approach ◽  
Continuous Variables ◽  
Integer Variables ◽  
System A ◽  
Global Optimization Methods

Abstract A global optimization approach for identifying the optimal product configuration and parameters is proposed to improve manufacturability measures including feasibility, cost, and time of production. Different product configurations, including alternative design candidates and production processes, are represented by an AND/OR graph. Product parameters are described by variables including continuous variables, integer variables, Boolean variables, and discrete variables. Two global optimization methods, genetic algorithm and simulated annealing, are employed for identifying the optimal product configuration and parameters. The introduced approach serves as a key component in an integrated concurrent design system. A case study example is given to show how the proposed method is used for solving the engineering problems.

Sign in / Sign up

Export Citation Format

Share Document