A New Hybrid Optimization Method Using Design Of Experiment Together With Artificial Neural Genetic Algorithm

Loading...
Thumbnail Image
Date
2011-10
Authors
Yahaya, Nor Zaiazmin
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Sains Malaysia
Abstract
In the engineering design process, it is a necessity to reduce the engineering design cycle time to meet the global market demand and also the customers need. Among the steps in the engineering design process, optimization process always consumed a lot of time and resources. This is because the optimization process involved a lot of parameters and infinite solutions that required a lot of experimental runs. A new a new hybrid optimization has been developed in this research that should be able to yield higher prediction accuracy for the optimal solution and at the same time requires only a minimum number of experimental runs without compromising the prediction accuracy. This new hybrid optimization method is developed by the integration of Design of Experiment (DOE), Artificial Neural Network (ANN) and Genetic Algorithm (GA). As a result of this research work, the new hybrid optimization method has outperformed the classical optimization method in average of 6.3% in terms of predicting the optimal input variables. Furthermore, the new hybrid optimization method reduced the number of experimental runs used to train the ANN, therefore reducing the overall total cost and shorten the engineering design cycle time.
Description
Keywords
A new hybrid optimization method using design , experiment together with artificial neural genetic algorithm
Citation