A neural-network-based methodology for the prediction of surface roughness in a turning process

Abstract

A neural-network-based methodology is proposed for predicting the surface roughness in a turning process by taking the acceleration of the radial vibration of the tool holder as feedback. Upper, most likely and lower estimates of the surface roughness are predicted by this method using very few experimental data for training and testing the network. The network model is trained using the back-propagation algorithm. The learning rate, the number of neurons in the hidden layer, the error goal, as well as the training and the testing dataset size, are found automatically in an adaptive manner. Since the training and testing data are collected from experiments, a data filtration scheme is employed to remove faulty data. The validation of the methodology is carried out for dry and wet turning of steel using high speed steel and carbide tools. It is observed that the present methodology is able to make accurate prediction of surface roughness by utilising small sized training and testing datasets.

A review of artificial intelligent approaches applied to part accuracy prediction

“…The MLP neural network with two hidden layers predicted the surface roughness with high accuracy (90%). Similarly, Kohli and Dixit (2005) developed MLP NN models for surface roughness prediction in dry and wet turning operations for high speed steel tools and carbide tools using as feed-back the radial vibration of the tool holder. The networks were trained by the back-propagation algorithm and the learning rate, the number of neurons in the hidden layer, and the training and testing dataset size were found automatically in an adaptive manner.…”

“…Different authors concluded that both logsig and tansig functions produce almost the same performance and it is not a critical parameter for modelling part accuracy (Kohli and Dixit, 2005;Zhong, Khoo and Han, 2006). However, other authors consider the mapping function selection an important factor for ANN models.…”

“…A methodology to predict the lower and upper bound of any non-linear function has been proposed by Ishibuchi and Tanaka (1991), which requires a simple modification of the back-propagation algorithm. This methodology was applied by Kohli and Dixit (2005) and Sonar, Dixit and Ojha (2006) to predict the lower and upper bound of a surface roughness prediction with MLP NN and RBF nets respectively.…”

Pre-evaluation on surface profile in turning process based on cutting parameters

“…The outputs were machined surface roughness, tangential force, axial force, spindle motor power, average flank wear, maximum flank wear, and nose wear. Kohli and Dixit et al [24] proposed a neural-network-based methodology for predicting the surface roughness in a turning process by taking the acceleration of the radial vibration of the tool holder as feedback. The network model was trained using BP algorithm, and its parameters were found automatically in an adaptive manner.…”

OPTIMIZATION OF CUTTING PARAMETERS ON MILD STEEL WITH HSS & CEMENTED CARBIDE TIPPED TOOLS USING ANN

“…However, factors such as geometry of cutting tool, tool wear, and joint material properties of both tool and work piece are uncontrollable [5]. One should develop techniques to evaluate the surface roughness of a product before machining in order to determine the required machining parameters such as feed rate, spindle speed and depth of cut for obtaining a desired surface roughness and product quality [6,7].…”