Predictive Modeling of Surface Roughness in MQL assisted Turning of SiC-Al Alloy Composites using Artificial Neural Network and Adaptive Neuro Fuzzy Inference System
Journal Title: Journal of Advanced Research in Manufacturing, Material Science & Metallurgical Engineering - Year 2018, Vol 5, Issue 3
Abstract
This research work deals with the analysis of surface roughness in turning SiC-Al Alloy composite through experimental investigation and predictive model formulation using Artificial neural network (ANN) and Adaptive neuro fuzzy inference system (ANFIS). The experiment has been carried out in MQL cutting condition where cutting speed, feed rate and depth of cut have been considered as input parameters to check the desired surface roughness response. Artificial neural network model with four different learning algorithms was used to predict surface roughness. Best performance was attained by 3-6-1 network structure with LM learning algorithm for training dataset and SCG learning algorithm for testing dataset. Higher correlation coefficient value of 0.99932 and 0.99888 prove the adequacy of the predicted ANN model. However, MAPE for ANN predicted value is 5.298 %. ANFIS model has been designed utilizing Gaussian membership function (Gaussmf) with 3 membership function for every input parameter and linear membership function for output parameter. For training both back propagation and hybrid method, 3 membership function were used. Based on the comparison of ANFIS with three types of membership function parameter training, hybrid method provides accurate results as it portrays MAPE value of 0.113542 when compared to the higher MAPE value of back propagation method. Surface viewer plot also suggested the effectiveness of hybrid model as it generates lower value of surface roughness. On the basis of mean absolute percentage error, it can be further concluded that hybrid method can make an impartially accurate prediction of surface roughness in comparison to back propagation method and artificial neural network which does prove the expediency of proposed hybrid model to reduce the surface roughness considerably.
Authors and Affiliations
Farhana Dilwar,
Keywords
Related Articles
- EP ID EP535314
- DOI -
- Views 121
- Downloads 0
Multi Objective Optimization of Pitting Corrosion Rate of Micro Plasma Arc Welded Titanium (Ti-6Al-4V) Alloy Sheets using TOPSIS
Titanium is having good anti-corrosion properties in both basic and acidic medium. However, when it is subjected to severe heating processing like welding, there is a possibility of corrosion formation at the weld joint....
Line Balancing Technique Improves Productivity At Chipset Tester Assembly Manufacturing: A Review
This paper reviews line balancing technique that used to for productivity purpose. Globalization is one of the key factors that brought growth in the chipset production demand. As a tester supplier for the chipset world,...
Experimental Study To Assess The Effect Of Electrode Material While Machining Inconel 718 Through Edm Process
Electrical discharge machining process is extensively used in tool and die making industry for processing of hardened materials with reasonable precision [1]. In the present work, comparison of MRR has been done using br...
Predictive Modeling of Surface Roughness in MQL assisted Turning of SiC-Al Alloy Composites using Artificial Neural Network and Adaptive Neuro Fuzzy Inference System
This research work deals with the analysis of surface roughness in turning SiC-Al Alloy composite through experimental investigation and predictive model formulation using Artificial neural network (ANN) and Adaptive neu...
Learning to Lean in a ‘Make in India’ Scenario
Make in India’ as a policy initiative is not only being watched by the world but is attracting good overseas support. There is a complementary opposition from local political groups, who await to see its debacle, though...