Modelling and Prediction of Surface Roughness and Power Consumption Using Parallel Extreme Learning Machine Based Particle Swarm Optimization

Ahmad Nooraziah, Tiagrajah V. Janahiraman

Research output: Book/ReportBook

Abstract

Prediction model allows the machinist to determine the values of the cutting performance before machining. Modelling using improved extreme learning machine based particle swarm optimization, IPSO-ELM has less parameters to adjust and also takes real number as particles while decreasing the norm of output weights and constraining the input weight and hidden biases within a reasonable range to improve the ELM performance. In order to solve the multi objectives modelling problem, we have proposed a parallel IPSO-ELM. In this research work, the best input weights and hidden biases for different performance were identified. The proposed method was able to model the training and the testing set with minimal error. The predicted result from the designed model was able to match the experimental data very closely.
Original languageEnglish
PublisherSpringer, Cham
Number of pages9
Volume2
DOIs
Publication statusPublished - 2014

Fingerprint

Particle swarm optimization (PSO)
Learning systems
Electric power utilization
Surface roughness
Machining
Testing

Cite this

@book{6b6860e1d6ad49b8b04abb64d7805f6d,
title = "Modelling and Prediction of Surface Roughness and Power Consumption Using Parallel Extreme Learning Machine Based Particle Swarm Optimization",
abstract = "Prediction model allows the machinist to determine the values of the cutting performance before machining. Modelling using improved extreme learning machine based particle swarm optimization, IPSO-ELM has less parameters to adjust and also takes real number as particles while decreasing the norm of output weights and constraining the input weight and hidden biases within a reasonable range to improve the ELM performance. In order to solve the multi objectives modelling problem, we have proposed a parallel IPSO-ELM. In this research work, the best input weights and hidden biases for different performance were identified. The proposed method was able to model the training and the testing set with minimal error. The predicted result from the designed model was able to match the experimental data very closely.",
author = "Ahmad Nooraziah and {V. Janahiraman}, Tiagrajah",
year = "2014",
doi = "10.1007/978-3-319-14066-7_31",
language = "English",
volume = "2",
publisher = "Springer, Cham",

}

TY - BOOK

T1 - Modelling and Prediction of Surface Roughness and Power Consumption Using Parallel Extreme Learning Machine Based Particle Swarm Optimization

AU - Nooraziah, Ahmad

AU - V. Janahiraman, Tiagrajah

PY - 2014

Y1 - 2014

N2 - Prediction model allows the machinist to determine the values of the cutting performance before machining. Modelling using improved extreme learning machine based particle swarm optimization, IPSO-ELM has less parameters to adjust and also takes real number as particles while decreasing the norm of output weights and constraining the input weight and hidden biases within a reasonable range to improve the ELM performance. In order to solve the multi objectives modelling problem, we have proposed a parallel IPSO-ELM. In this research work, the best input weights and hidden biases for different performance were identified. The proposed method was able to model the training and the testing set with minimal error. The predicted result from the designed model was able to match the experimental data very closely.

AB - Prediction model allows the machinist to determine the values of the cutting performance before machining. Modelling using improved extreme learning machine based particle swarm optimization, IPSO-ELM has less parameters to adjust and also takes real number as particles while decreasing the norm of output weights and constraining the input weight and hidden biases within a reasonable range to improve the ELM performance. In order to solve the multi objectives modelling problem, we have proposed a parallel IPSO-ELM. In this research work, the best input weights and hidden biases for different performance were identified. The proposed method was able to model the training and the testing set with minimal error. The predicted result from the designed model was able to match the experimental data very closely.

UR - https://link.springer.com/chapter/10.1007/978-3-319-14066-7_31#citeas

U2 - 10.1007/978-3-319-14066-7_31

DO - 10.1007/978-3-319-14066-7_31

M3 - Book

VL - 2

BT - Modelling and Prediction of Surface Roughness and Power Consumption Using Parallel Extreme Learning Machine Based Particle Swarm Optimization

PB - Springer, Cham

ER -