Open AccessModel Development for Prediction of Surface Roughness by using of AI Technique
Author(s) -
Mohd. Tauseef,
Dheeraj Kumar Verma
Publication year - 2020
Publication title -
international journal of scientific research in science, engineering and technology
Language(s) - English
Resource type - Journals
eISSN - 2395-1990
pISSN - 2394-4099
DOI - 10.32628/ijsrset207636
Subject(s) - artificial neural network , mean squared error , computer science , surface roughness , matlab , machine learning , regression , slicing , artificial intelligence , toolbox , set (abstract data type) , function (biology) , backpropagation , regression analysis , feedforward neural network , statistics , mathematics , physics , quantum mechanics , evolutionary biology , world wide web , biology , programming language , operating system
The surface roughness of manufactured product is final results of the turning technique parameters, and an critical characteristics that outline product first-rate, aesthetics etc. It imposes one of the most essential constraints for the choice of machines and slicing parameters in manner planning. In this paper, Artificial Neural Network (ANN) method has been used to develop surface roughness prediction model the use of experimental statistics, wherein Feed Forward Neural Network (FFNN) the usage of Back Propagation set of rules and Levenberg-Marquardt education function has been used. The work has been done using Neural etwork Toolbox in MATLAB. The overall performance of the version has been assessed based totally on Regression analysis, Mean Square Error (MSE) and Magnitude of Relative Error (MRE). A three-2-1 model with two neurons in the hidden layer turned into discovered to be the excellent developed model, having universal regression ( R) cost of zero.9923 and pleasant validation overall performance MSE value of 0.00913. The ANN model confirmed incredible consequences for forecasting