{"title":"基于机器学习算法的电火花加工参数高级优化:形状记忆合金的实验研究","authors":"Ranjit Singh, Ravi Pratap Singh, Rajeev Trehan","doi":"10.1016/j.sintl.2022.100179","DOIUrl":null,"url":null,"abstract":"<div><p>In the era of advancement and progressive fourth industrial revolution culture, the demand of advanced and smart engineering materials has been increased. In this way, shape memory alloys are an excellent choice for industrial applications such as orthopedic implacers, actuators, micro-tools, fitting and screening elements, aircraft component components, military instruments, fabricating elements, and bio-medical devices, among others. This paper has been aimed to attempt the machine learning (ML) algorithms-based optimization of the different process inputs in electrical discharge machining of Cu-based shape memory alloy. The current study focused on study the behavior of response parameters along with the variation in machining input parameters The considered process input factors are namely as; pulse on time (Ton), pulse off time (Toff), peak current (Ip), and gap voltage (GV) and their effects were studied on dimensional deviation (DD) and tool wear rate (TWR). The central composite design matrix has been employed for planning the main runs. The 2-D and 3-D graphs represents the behavior of the response parameters along with variations in the machining inputs. The novelty of the work is machining of Cu-based Shape Memory Alloy (SMA) in EDM operations and optimization of parameters using Machine Learning techniques. Furthermore, machine learning based, single and multi-objective optimization of investigated responses were conducted using the desirability approach, Genetic Algorithm (GA) and Teacher Learning based Optimization (TLBO) techniques. The parametric combination attained for optimization of multiple responses (TWR and DD) is: Ton = 90.10 μs, Toff = 149.69 μs, Ip = 24.59 A & GV = 60 V; Ton = 255 μs, Toff = 15 μs, Ip = 50 A & GV = 15 V; Ton = 255 μs, Toff = 15 μs, Ip = 50 A & GV = 15 V, using desirability approach, GA method and TLBO method, respectively.</p></div>","PeriodicalId":21733,"journal":{"name":"Sensors International","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666351122000249/pdfft?md5=36969b635e3b41ed051f61e0fb03d946&pid=1-s2.0-S2666351122000249-main.pdf","citationCount":"3","resultStr":"{\"title\":\"Machine learning algorithms based advanced optimization of EDM parameters: An experimental investigation into shape memory alloys\",\"authors\":\"Ranjit Singh, Ravi Pratap Singh, Rajeev Trehan\",\"doi\":\"10.1016/j.sintl.2022.100179\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><p>In the era of advancement and progressive fourth industrial revolution culture, the demand of advanced and smart engineering materials has been increased. In this way, shape memory alloys are an excellent choice for industrial applications such as orthopedic implacers, actuators, micro-tools, fitting and screening elements, aircraft component components, military instruments, fabricating elements, and bio-medical devices, among others. This paper has been aimed to attempt the machine learning (ML) algorithms-based optimization of the different process inputs in electrical discharge machining of Cu-based shape memory alloy. The current study focused on study the behavior of response parameters along with the variation in machining input parameters The considered process input factors are namely as; pulse on time (Ton), pulse off time (Toff), peak current (Ip), and gap voltage (GV) and their effects were studied on dimensional deviation (DD) and tool wear rate (TWR). The central composite design matrix has been employed for planning the main runs. The 2-D and 3-D graphs represents the behavior of the response parameters along with variations in the machining inputs. The novelty of the work is machining of Cu-based Shape Memory Alloy (SMA) in EDM operations and optimization of parameters using Machine Learning techniques. Furthermore, machine learning based, single and multi-objective optimization of investigated responses were conducted using the desirability approach, Genetic Algorithm (GA) and Teacher Learning based Optimization (TLBO) techniques. The parametric combination attained for optimization of multiple responses (TWR and DD) is: Ton = 90.10 μs, Toff = 149.69 μs, Ip = 24.59 A & GV = 60 V; Ton = 255 μs, Toff = 15 μs, Ip = 50 A & GV = 15 V; Ton = 255 μs, Toff = 15 μs, Ip = 50 A & GV = 15 V, using desirability approach, GA method and TLBO method, respectively.</p></div>\",\"PeriodicalId\":21733,\"journal\":{\"name\":\"Sensors International\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2022-01-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.sciencedirect.com/science/article/pii/S2666351122000249/pdfft?md5=36969b635e3b41ed051f61e0fb03d946&pid=1-s2.0-S2666351122000249-main.pdf\",\"citationCount\":\"3\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Sensors International\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S2666351122000249\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Sensors International","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2666351122000249","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 3
摘要
在第四次工业革命文化不断进步的时代,对先进和智能工程材料的需求不断增加。通过这种方式,形状记忆合金是工业应用的绝佳选择,例如骨科植入物,执行器,微型工具,配件和筛选元件,飞机部件部件,军事仪器,制造元件和生物医疗设备等。本文旨在尝试基于机器学习算法的铜基形状记忆合金电火花加工中不同工艺输入的优化。目前的研究重点是研究响应参数随加工输入参数变化的行为,考虑的工艺输入因素为;研究了脉冲开启时间(Ton)、脉冲关闭时间(Toff)、峰值电流(Ip)和间隙电压(GV)及其对刀具尺寸偏差(DD)和磨损率(TWR)的影响。中心复合设计矩阵已被用于规划主要运行。二维和三维图形表示响应参数随加工输入变化的行为。这项工作的新颖之处在于在电火花加工中加工铜基形状记忆合金(SMA),并使用机器学习技术优化参数。此外,利用可取性方法、遗传算法(GA)和基于教师学习的优化(TLBO)技术,对调查结果进行了基于机器学习的单目标和多目标优化。对多重响应(TWR和DD)优化得到的参数组合为:Ton = 90.10 μs, Toff = 149.69 μs, Ip = 24.59 A &gv = 60v;吨= 255μs,有钱人= 15μs, Ip = 50,gv = 15 v;吨= 255μs,有钱人= 15μs, Ip = 50,GV = 15 V,分别采用可取性法、GA法和TLBO法。
Machine learning algorithms based advanced optimization of EDM parameters: An experimental investigation into shape memory alloys
In the era of advancement and progressive fourth industrial revolution culture, the demand of advanced and smart engineering materials has been increased. In this way, shape memory alloys are an excellent choice for industrial applications such as orthopedic implacers, actuators, micro-tools, fitting and screening elements, aircraft component components, military instruments, fabricating elements, and bio-medical devices, among others. This paper has been aimed to attempt the machine learning (ML) algorithms-based optimization of the different process inputs in electrical discharge machining of Cu-based shape memory alloy. The current study focused on study the behavior of response parameters along with the variation in machining input parameters The considered process input factors are namely as; pulse on time (Ton), pulse off time (Toff), peak current (Ip), and gap voltage (GV) and their effects were studied on dimensional deviation (DD) and tool wear rate (TWR). The central composite design matrix has been employed for planning the main runs. The 2-D and 3-D graphs represents the behavior of the response parameters along with variations in the machining inputs. The novelty of the work is machining of Cu-based Shape Memory Alloy (SMA) in EDM operations and optimization of parameters using Machine Learning techniques. Furthermore, machine learning based, single and multi-objective optimization of investigated responses were conducted using the desirability approach, Genetic Algorithm (GA) and Teacher Learning based Optimization (TLBO) techniques. The parametric combination attained for optimization of multiple responses (TWR and DD) is: Ton = 90.10 μs, Toff = 149.69 μs, Ip = 24.59 A & GV = 60 V; Ton = 255 μs, Toff = 15 μs, Ip = 50 A & GV = 15 V; Ton = 255 μs, Toff = 15 μs, Ip = 50 A & GV = 15 V, using desirability approach, GA method and TLBO method, respectively.