ISSN Print: 2381-1072  ISSN Online: 2381-1080
Engineering and Technology  
Manuscript Information
 
 
Development of Probabilistic Neural Network Model for Weld Quality Prediction
Engineering and Technology
Vol.5 , No. 2, Publication Date: May 30, 2018, Page: 21-27
742 Views Since May 30, 2018, 550 Downloads Since May 30, 2018
 
 
Authors
 
[1]    

John Edwin Raja Dhas, Department of Automobile Engineering, Noorul Islam Centre for Higher Education, Kumaracoil, Tamil Nadu, India.

[2]    

Moni Satheesh, Department of Automobile Engineering, Noorul Islam Centre for Higher Education, Kumaracoil, Tamil Nadu, India.

 
Abstract
 

There has been a great increase in the number of automatic and semi-automatic processes designed to speed up welding production. Submerged Arc Welding (SAW) is often the method of choice in pressure vessel fabrication. This process features high production rates, welding energy and/or welding speed and requires minimal operator skill. The selection of appropriate parameters in SAW is essential, not only to optimize the welding process in order to maintain the highest level of productivity, but also to obtain the most desirable mechanical properties of the weld. This paper explores the development of Probabilistic Neural Network (PNN) model for weld quality prediction. Experiments were designed and conducted by Taguchi’s design of experiments. Results from the PNN model are compared with the Neural Network model trained with Back Propagation (BPNN) algorithm in terms of computational speed and accuracy. The performance of PNN is better than BPNN model and is reasonably more accurate. Confirmatory experiments are done to validate this approach and reported.


Keywords
 

Weld Parameters, Quality, Taguchi Method, Probabilistic Neural Network


Reference
 
[01]    

Welding Handbook, vol. 2, second ed., American Welding Society, 1978.

[02]    

Yang, L. J., Chandel, R. S. and Bibby, M. J., An analysis of curvilinear regression equations for modeling the submerged-arc welding process, Journal of Materials Processing and Technology 37 (1993) 601-611.

[03]    

Gunaraj, V. and Murgun, N. Application of response surface methodology for predicting weld bead quality in submerged arc welding of pipes. Journal of Materials Processing Technology 88 (1999) 266-275.

[04]    

Markelj, F. and Tusek, J. Algorithmic optimization of parameters in tungsten inert gas welding of stainless-steel sheet, Science and Technology of Welding and Joining 6 (2001) 375-382.

[05]    

Allen, T. T., Richardson, R. W., Tagliabue, D. P. and Maoul, G. P. Statistical process design for robotic GMA welding of sheet metal, Welding Journal Supplement (2002) 69-70.

[06]    

Kim, I. S., Son, J. S., Kim, I. G., Kim, J. Y., Kim and O. S. A study on relationship between process variables and bead penetration for robotic CO2 arc welding. Journal of Materials Processing Technology 136 (2003) 139-145.

[07]    

Murugan, N. and Gunaraj, V. Prediction and control of weld bead geometry and shape relationships in submerged arc welding of pipes, Journal of Materials Processing Technology 168 (2005) 478-487.

[08]    

Edwin Raja Dhas J and Satheesh M. Multiple objective optimization of Submerged Arc Weld process parameters using Grey based Taguchi method, International Journal of Industrial and Systems Engineering 12 (2012) 331-345.

[09]    

Satheesh, M. and Edwin Raja Dhas, J. Multi objective optimization of weld parameters of boiler steel using fuzzy based desirability function. Journal of Engineering Science and Technology Review 7 (2014) 29-36.

[10]    

Edwin Raja Dhas J. and Satheesh M. Multi Objective Optimization of FCAW parameters using Grey based Taguchi with entropy Technique, International Journal of Industrial and Systems Engineering 19 (2015) 190-205.

[11]    

Edwin Raja Dhas, J., Somasundaram Kumanan. and Jesuthanam, C. P. Prediction of Weld Quality using intelligent decision making tools. International Journal of Artificial Intelligence Research. 1 (2012) 131-148.

[12]    

Anand K., Birendra Kumar Barik, K. Tamilmannan, P. Sathiya Artificial neural network modeling studies to predict the friction welding process parameters of Incoloy 800H joints, Engineering Science and Technology, an International Journal 18 (2015) 394-407.

[13]    

Edwin Raja Dhas J., Stalin R. S. and Rajeesh J. RBF neural network model for machining quality prediction International Journal of Modeling Identification and Control 20 (2013) 174-180.

[14]    

Specht, D. F., 1990. Probabilistic neural networks. Neural Networks 1 (3), 109-118.

[15]    

Berthold, M. and Diamond, J. Constructive training of probabilistic neural networks. Neurocomputing (1998) 167-183.

[16]    

Specht, D. F. and Romsdahl, H. Experience with adaptive probabilistic neural network and adaptive general regression neural network. In. Proceedings of the IEEE International Conference on Neural Networks 2 (1994) 1203-1208.

[17]    

Parzen, E., On the estimation of a probability density function and mode. Annuals of Mathematical Statistics. 3 (1962) 1065-1076.

[18]    

Tan, A. C. and Gilbert, D. An empirical comparison of supervised machine learning techniques in bioinformatics. Proceedings of the 1st Asia Pacific Bioinformatics Conference (2003) 219-222.

[19]    

Edwin Raja Dhas J and Somasundaram Kumanan Evolutionary SVM modeling of weld residual stress, Applied Soft Computing, Vol 26, (2016) 474-482.

[20]    

Ramanan G, Edwin Raja Dhas J and Jai Aultrin K S Multi Response Prediction of Machining Process Parameters Using Artificial Neural Network International Journal of Mechanical Engineering and Technology, Volume 8, Issue 5, (2017) 866-876.

[21]    

G. Ramanan, J. Edwin Raja Dhas, Neural Network Prediction and Analysis of Material Removal Process during Wire Cut Electrical Discharge Machining, REST Journal on Emerging trends in Modelling and Manufacturing 3 (1) (2017) 7-11.





 
  Join Us
 
  Join as Reviewer
 
  Join Editorial Board
 
share:
 
 
Submission
 
 
Membership