TY - CHAP
T1 - Application of Hybrid ANN and PSO for Prediction of Surface Roughness Under Biodegradable Nano-lubricant
AU - Okokpujie, Imhade P.
AU - Tartibu, Lagouge K.
N1 - Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2023
Y1 - 2023
N2 - Modelling of surface roughness is challenging because machining conditions cannot be well covered by theoretical models. In this section, 50 configurations corresponding to various spindle speed, feed rate, length of cut, depth of cut, helix angle were considered to predict the surface roughness of the end milling machining of an AA8112 alloy. In this study, an ANN-PSO approach has been considered for the development of a suitable prediction model for surface roughness. The sensitivity analysis of numerous Particle Swarm Optimization (PSO) parameters, including the population size of the swarm, the number of neurons in the hidden layer, and the magnitude of the acceleration factors, was carried out. This study reveals that the prediction performance metric is closely related to the model configuration or parameters. The best configurations of the ANN-PSO models were identified. This study shows that the hybrid ANN-PSO enhances the performance of ANN. The proposed approach would address time-consuming and expensive experiment required to identify a configuration that yield the minimum surface roughness.
AB - Modelling of surface roughness is challenging because machining conditions cannot be well covered by theoretical models. In this section, 50 configurations corresponding to various spindle speed, feed rate, length of cut, depth of cut, helix angle were considered to predict the surface roughness of the end milling machining of an AA8112 alloy. In this study, an ANN-PSO approach has been considered for the development of a suitable prediction model for surface roughness. The sensitivity analysis of numerous Particle Swarm Optimization (PSO) parameters, including the population size of the swarm, the number of neurons in the hidden layer, and the magnitude of the acceleration factors, was carried out. This study reveals that the prediction performance metric is closely related to the model configuration or parameters. The best configurations of the ANN-PSO models were identified. This study shows that the hybrid ANN-PSO enhances the performance of ANN. The proposed approach would address time-consuming and expensive experiment required to identify a configuration that yield the minimum surface roughness.
KW - ANN
KW - ANN-PSO models
KW - Machining
KW - Particle swarm optimization
KW - Surface roughness
UR - http://www.scopus.com/inward/record.url?scp=85165992477&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-35455-7_12
DO - 10.1007/978-3-031-35455-7_12
M3 - Chapter
AN - SCOPUS:85165992477
T3 - Studies in Systems, Decision and Control
SP - 263
EP - 288
BT - Studies in Systems, Decision and Control
PB - Springer Science and Business Media Deutschland GmbH
ER -