《机电工程》杂志社科技期刊核心期刊论文投稿机电一体化机械、仪表电工自动化、计算机浙江大学主办

Introduction

International Standard Serial Number:

ISSN 1001-4551

Sponsor:

Zhejiang University;

Zhejiang Machinery and Electrical Group

Edited by:

Editorial of Journal of Mechanical & Electrical Engineering

Chief Editor:

ZHAO Qun

Vice Chief Editor:

TANG ren-zhong,

LUO Xiang-yang

Tel:

86-571-87041360,87239525

Fax:

86-571-87239571

Add:

No.9 Gaoguannong,Daxue Road,Hangzhou,China

P.C:

310009

E-mail:

meem_contribute@163.com

RUL prediction based on EG-SSMA-DELM milling machine tool

Published:2023-10-30 author:ZHANG Tianxiao, GU Yanling, AN Wenjie. Browse: 291 Check PDF documents

RUL prediction based on EG-SSMA-DELM milling machine tool

ZHANG Tianxiao, GU Yanling, AN Wenjie

(School of Mechanical Engineering, Shenyang University of Technology, Shenyang 110870, China)

Abstract: Aiming at the problems of workpiece scrap and key component damage caused by tool failure, a tool wear prediction model based on elite opposition-based learning golden sine slime mould algorithm-deep extreme learning machine (EG-SSMA-DELM)was proposed. Firstly, in the slime mould algorithm (SMA), the elite opposition-based learning (EOBL) and golden sine algorithms (GSA) were used to optimize the initial slime mould population, the diversity of the initial population, the update method of the initial SMA search individual position, and the convergence speed and global search ability of the algorithm were improved to obtain the optimal parameters. Then, the improved SMA algorithm was used to jointly optimize the bias and input weight of the encoder in deep extreme learning machine (DELM), define different numbers of hidden layer neurons, and the ReLU activation function was used to arrange the parameters of DELM ideally. Finally, the projected features were inputted into DELM for training and prediction according to the optimal parameters, so as to predict the remaining service life of the tool. The research results show that compared with the classical deep limit learning machine method, the proposed EG-SSMA-DELM method reduces the root mean square error (RMSE) by 19.60% on average,and improves the prediction accuracy by 16.00%. Comparing with other deep learning algorithms, this algorithm model has better feasibility, monotonicity and stronger robustness, and has certain application value for the study of the remaining life of actual engineering tool wear. The algorithm model has certain application value for the study of tool wear residual life in practical engineering.

Key words: remaining useful life(RUL); tool life prediction; elite oppositionbased learning(EOBL); golden sine algorithm(GSA); slime mould algorithm(SMA); deep extreme learning machine(DELM)

Advanced Search

Online Office System

Chinese Core Periodicals
Chinese Sci-tech Core Periodicals
SA, INSPEC Indexed
CSA: T Indexed
UPD:Indexed