www.tjprc.org editor@tjprc.org
International Journal of Mechanical and Production
Engineering Research and Development (IJMPERD)
ISSN(P): 2249-6890; ISSN(E): 2249-8001
Vol. 4, Issue 6, Dec 2014, 13-18
© TJPRC Pvt. Ltd.
MULTI-OBJECTIVE OPTIMIZATION OF CUTTING PARAMETERS IN HARD TURNING
PROCESS USING GENETIC ALGORITHM (GA) & ARTIFICIAL NEURAL
NETWORK(ANN)
A. V. N. L. SHARMA
1
& K. VENKATA SUBBAIAH
2
1
Department of Mechanical Engineering, BITS, Visakhapatnam, Andhra Pradesh, India
2
Department of Mechanical Engineering, Andhra University, Visakhapatnam, Andhra Pradesh, India
ABSTRACT
Manufacturing industry today faces the challenge of having to develop high quality products faster and
economically than ever before. Therefore optimization is seen as an innovative technique under certain premises.
Optimization in turning means determination of the optimal set of machining parameters to satisfy the objectives within the
operational constraints. Predictive modeling is essential for understanding and optimization of the machining process.
The aim of this study is to develop an integrated model to optimize the cutting parameters that are affecting the quality of
surface produced in hard turning process on EN 353 metal. Three input parameters were selected for study: cutting speed,
feed & depth of cut to determine the optimal cutting parameters required for minimum surface roughness, power
consumption and for maximum metal removal rate. Mathematical equations are formulated as objective functions, to
determine the optimal cutting parameters so that minimization of surface roughness/power consumption and maximization
of metal removal rate are evaluated by using MATLAB software. Genetic Algorithm(GA) supported with tested ANN is
utilized to determine the best combinations of cutting parameters through optimization process. Artificial Neural Network
(ANN) on back propagation learning with hidden neurons is used to validate the model The trained machined data was
tested and the results show that the model has the ability to solve many problems including predicting, modeling and
measuring experimental knowledge under dry environment. From these results, it can be easily realized that the developed
study is reliable and suitable for solving the other parameters encountered in metal cutting operations as the same as
surface roughness.
KEYWORDS: Artificial Neural Network (ANN), Genetic Algorithm(GA), Regression Analysis, CVD
INTRODUCTION
Manufacturing of goods involves the machining of parts or components with the cutting tools set or loaded on the
machine tools. The selection of suitable material and the suitable cutting tool combination plays greater role in
Production / Manufacturing parts or components in industry. The turning process parameter optimization is highly
constrained and non – linear. ANNs have been trained based on back propagation(BP) learning algorithm and tested to
control the performance of trained ANN model. By adapting the tested ANN model with powerful GA optimization was
applied to achieve best combination of cutting parameters. The performance of ANN is carried out with experimental data
of CVD on EN 353. The simulation is carried out for 3 parameters over surface roughness, MRR and power consumption.
The input-output dataset consisting of 27 patterns was divided randomly in two categories: training dataset consist of 75%
of the data and test dataset which consist 25 % of the data. There are 20 training patterns considered for ANN modeling of