An Approach to optimize ANN Meta model with Multi Objective Genetic Algorithm for Multi-disciplinary Shape optimization
Ram Krishna Rathore1, Amit Sarda2, Rituraj Chandrakar3

1Ram Krishna Rathore, Mechanical engineering Department, Christian College of Engineering & Technology, Bhilai (C.G), India.
2Amit Sarda, Mechanical engineering Department, CCET, Bhilai, India.
3Rituraj Chandrakar, Mechanical engineering Department, CSIT, Durg, India.

Manuscript received on February 15, 2012. | Revised Manuscript received on February 20, 2012. | Manuscript published on March 05, 2012. | PP: 200-207 | Volume-2 Issue-1, March 2012. | Retrieval Number: A0414022112/2012©BEIESP
Open Access | Ethics and Policies | Cite 
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (

Abstract: In several design cases, designers need to optimize a number of responses concurrently. A general approach for the multiple response cases optimization start with using the regression models to calculate the correlations between response functions and control factors. Then, a system for collecting various response functions together into a one quantity, such as an objective function, is engaged and, at last, an optimization technique is used to calculate the best combinations for the control functions. A different method proposed in this paper is to use an artificial neural network (ANN) to calculate the parameter response functions. At the optimization stage, a multi objective genetic algorithm (MOGA) is used in combination with an objective functions to establish the optimum conditions for the control functions. A crane hook example has been taken to optimize multiple shape parameter responses to with stand a new loading condition. The results estimate the reduction in mass and sufficient factor of safety to show the proposed approach for the optimization of multi- disciplinary shape optimization problems.

Keywords: ANN, MOGA, Shape optimization, Meta modeling