Paolo Di Barba, Fabrizio Dughiero, Michele Forzan, David A. Lowther, Antonio Marconi, Maria Evelina Mognaschi, Jan K. Sykulski
{"title":"感应加热过程的神经元模型和迁移学习(TEAM 36题)","authors":"Paolo Di Barba, Fabrizio Dughiero, Michele Forzan, David A. Lowther, Antonio Marconi, Maria Evelina Mognaschi, Jan K. Sykulski","doi":"10.3233/jae-230087","DOIUrl":null,"url":null,"abstract":"The authors explore the possibility of applying a convolutional Naeural Network (CNN) to the solution of coupled electromagnetic and thermal problem, focusing on the classical problem of induction heating systems, traditionally solved by resorting to Finite Element (FE) models. In fact, FE modelling is widely used in the design of induction heating systems due its accuracy, even if the solution of a coupled nonlinear problem is expensive in terms of computational time and hardware resources, notably in 3D analysis. A model based on CNN could be an interesting alternative; in fact, CNN is a learning model selected for its excellent ability to converge, even when trained with a limited dataset. CNNs are able to treat images as input and they are used here as follows: given a temperature map in the workpiece, identify the corresponding vector of current, frequency and process heating time; this mapping is a model of the inverse induction heating problem. Specifically, we consider as an example the induction heating of a cylindrical steel billet, made of C45 steel, placed in a solenoidal inductor coil exhibiting the same axial length of the billet (TEAM 36 problem). A thorough heating process is usually applied before hot working of the billet, as in an extrusion process, but this methodology can be applied also in the design of induction hardening processes. First, a CNN has been trained from scratch by means of a dataset of FE solutions of coupled electromagnetic and thermal problems. For the sake of a comparison, a transfer learning technique is applied using GoogLeNet, i.e. a Deep Convolutional Neural Network able to classify images: starting from the pre-trained GoogLeNet, its training has been subsequently refined with the dataset of solutions from FE analyses. When the training dataset contains a limited number of samples, GoogleNet shows good accuracy in predicting the process parameters; in the case of a high number of samples in the training set, namely beyond a threshold like e.g. 1500, both CNNs show good accuracy of the result.","PeriodicalId":50340,"journal":{"name":"International Journal of Applied Electromagnetics and Mechanics","volume":"69 4","pages":""},"PeriodicalIF":1.1000,"publicationDate":"2023-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Neural metamodels and transfer learning for induction heating processes (TEAM 36 problem)\",\"authors\":\"Paolo Di Barba, Fabrizio Dughiero, Michele Forzan, David A. Lowther, Antonio Marconi, Maria Evelina Mognaschi, Jan K. Sykulski\",\"doi\":\"10.3233/jae-230087\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The authors explore the possibility of applying a convolutional Naeural Network (CNN) to the solution of coupled electromagnetic and thermal problem, focusing on the classical problem of induction heating systems, traditionally solved by resorting to Finite Element (FE) models. In fact, FE modelling is widely used in the design of induction heating systems due its accuracy, even if the solution of a coupled nonlinear problem is expensive in terms of computational time and hardware resources, notably in 3D analysis. A model based on CNN could be an interesting alternative; in fact, CNN is a learning model selected for its excellent ability to converge, even when trained with a limited dataset. CNNs are able to treat images as input and they are used here as follows: given a temperature map in the workpiece, identify the corresponding vector of current, frequency and process heating time; this mapping is a model of the inverse induction heating problem. Specifically, we consider as an example the induction heating of a cylindrical steel billet, made of C45 steel, placed in a solenoidal inductor coil exhibiting the same axial length of the billet (TEAM 36 problem). A thorough heating process is usually applied before hot working of the billet, as in an extrusion process, but this methodology can be applied also in the design of induction hardening processes. First, a CNN has been trained from scratch by means of a dataset of FE solutions of coupled electromagnetic and thermal problems. For the sake of a comparison, a transfer learning technique is applied using GoogLeNet, i.e. a Deep Convolutional Neural Network able to classify images: starting from the pre-trained GoogLeNet, its training has been subsequently refined with the dataset of solutions from FE analyses. When the training dataset contains a limited number of samples, GoogleNet shows good accuracy in predicting the process parameters; in the case of a high number of samples in the training set, namely beyond a threshold like e.g. 1500, both CNNs show good accuracy of the result.\",\"PeriodicalId\":50340,\"journal\":{\"name\":\"International Journal of Applied Electromagnetics and Mechanics\",\"volume\":\"69 4\",\"pages\":\"\"},\"PeriodicalIF\":1.1000,\"publicationDate\":\"2023-11-21\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"International Journal of Applied Electromagnetics and Mechanics\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://doi.org/10.3233/jae-230087\",\"RegionNum\":4,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"ENGINEERING, ELECTRICAL & ELECTRONIC\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Journal of Applied Electromagnetics and Mechanics","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.3233/jae-230087","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
Neural metamodels and transfer learning for induction heating processes (TEAM 36 problem)
The authors explore the possibility of applying a convolutional Naeural Network (CNN) to the solution of coupled electromagnetic and thermal problem, focusing on the classical problem of induction heating systems, traditionally solved by resorting to Finite Element (FE) models. In fact, FE modelling is widely used in the design of induction heating systems due its accuracy, even if the solution of a coupled nonlinear problem is expensive in terms of computational time and hardware resources, notably in 3D analysis. A model based on CNN could be an interesting alternative; in fact, CNN is a learning model selected for its excellent ability to converge, even when trained with a limited dataset. CNNs are able to treat images as input and they are used here as follows: given a temperature map in the workpiece, identify the corresponding vector of current, frequency and process heating time; this mapping is a model of the inverse induction heating problem. Specifically, we consider as an example the induction heating of a cylindrical steel billet, made of C45 steel, placed in a solenoidal inductor coil exhibiting the same axial length of the billet (TEAM 36 problem). A thorough heating process is usually applied before hot working of the billet, as in an extrusion process, but this methodology can be applied also in the design of induction hardening processes. First, a CNN has been trained from scratch by means of a dataset of FE solutions of coupled electromagnetic and thermal problems. For the sake of a comparison, a transfer learning technique is applied using GoogLeNet, i.e. a Deep Convolutional Neural Network able to classify images: starting from the pre-trained GoogLeNet, its training has been subsequently refined with the dataset of solutions from FE analyses. When the training dataset contains a limited number of samples, GoogleNet shows good accuracy in predicting the process parameters; in the case of a high number of samples in the training set, namely beyond a threshold like e.g. 1500, both CNNs show good accuracy of the result.
期刊介绍:
The aim of the International Journal of Applied Electromagnetics and Mechanics is to contribute to intersciences coupling applied electromagnetics, mechanics and materials. The journal also intends to stimulate the further development of current technology in industry. The main subjects covered by the journal are:
Physics and mechanics of electromagnetic materials and devices
Computational electromagnetics in materials and devices
Applications of electromagnetic fields and materials
The three interrelated key subjects – electromagnetics, mechanics and materials - include the following aspects: electromagnetic NDE, electromagnetic machines and devices, electromagnetic materials and structures, electromagnetic fluids, magnetoelastic effects and magnetosolid mechanics, magnetic levitations, electromagnetic propulsion, bioelectromagnetics, and inverse problems in electromagnetics.
The editorial policy is to combine information and experience from both the latest high technology fields and as well as the well-established technologies within applied electromagnetics.