Pub Date : 2022-06-29DOI: 10.36001/phme.2022.v7i1.3346
Jin-Woong Kang, C. Varnier, A. Mosallam, N. Zerhouni, Fares Ben Youssef, Nannan Shen
In the area of well construction, the tool reliability and the field environment are two contributing factors that influence drilling job efficiency and success. Either using high specification tools in low-risk environmental or applying tools of low reliability in harsh environments is inadvisable. Thus, how to select a suitable tool fitting the environment of an approaching drilling job is of great significance for tool planning. However, today, the tool selection decision is not optimized because it is often based on partial data availability and understanding. This paper presents an indicator called tool compatibility index, which can support improved tool selection decision making. This index takes part reliability, part criticality, and field environment into consideration, and gives a score indicating the compatibility of the tool to a specific environment. Moreover, the tool compatibility index is computed based on a weighted average method, which is computation simple and can be easily deployed. This work is part of a long-term project aiming to construct a risk-based decision advisor for drilling and measurement tools.
{"title":"Tool Compatibility Index: Indicator Enables Improved Tool Selection for Well Construction","authors":"Jin-Woong Kang, C. Varnier, A. Mosallam, N. Zerhouni, Fares Ben Youssef, Nannan Shen","doi":"10.36001/phme.2022.v7i1.3346","DOIUrl":"https://doi.org/10.36001/phme.2022.v7i1.3346","url":null,"abstract":"In the area of well construction, the tool reliability and the field environment are two contributing factors that influence drilling job efficiency and success. Either using high specification tools in low-risk environmental or applying tools of low reliability in harsh environments is inadvisable. Thus, how to select a suitable tool fitting the environment of an approaching drilling job is of great significance for tool planning. However, today, the tool selection decision is not optimized because it is often based on partial data availability and understanding.\u0000This paper presents an indicator called tool compatibility index, which can support improved tool selection decision making. This index takes part reliability, part criticality, and field environment into consideration, and gives a score indicating the compatibility of the tool to a specific environment. Moreover, the tool compatibility index is computed based on a weighted average method, which is computation simple and can be easily deployed. This work is part of a long-term project aiming to construct a risk-based decision advisor for drilling and measurement tools.","PeriodicalId":422825,"journal":{"name":"PHM Society European Conference","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126202120","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-29DOI: 10.36001/phme.2022.v7i1.3294
Van-Thai Nguyen, P. Do, A. Voisin
Maintenance planning for complex systems has still been a challenging problem. Firstly, integrating multiple dependency types into maintenance models makes them more realistic, however, more complicated to solve and analyze. Secondly, the number of maintenance decision variables needed to be optimized increases rapidly in the number of components, causing computational expensive for optimization algorithms. To face these issues, this thesis aims to incorporate multiple kinds of dependencies into maintenance models as well as to take advantage of recent advances in artificial intelligence field to effectively optimize maintenance polices for large-scale multi-component systems.
{"title":"Artificial-Intelligence-Based Maintenance Scheduling for Complex Systems with Multiple Dependencies","authors":"Van-Thai Nguyen, P. Do, A. Voisin","doi":"10.36001/phme.2022.v7i1.3294","DOIUrl":"https://doi.org/10.36001/phme.2022.v7i1.3294","url":null,"abstract":"Maintenance planning for complex systems has still been a challenging problem. Firstly, integrating multiple dependency types into maintenance models makes them more realistic, however, more complicated to solve and analyze. Secondly, the number of maintenance decision variables needed to be optimized increases rapidly in the number of components, causing computational expensive for optimization algorithms. To face these issues, this thesis aims to incorporate multiple kinds of dependencies into maintenance models as well as to take advantage of recent advances in artificial intelligence field to effectively optimize maintenance polices for large-scale multi-component systems.","PeriodicalId":422825,"journal":{"name":"PHM Society European Conference","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127630701","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-29DOI: 10.36001/phme.2022.v7i1.3324
A. Kamtsiuris, F. Raddatz, Gerko Wende
In the field of Maintenance, Repair and Overhaul (MRO), stakeholders such as operators or service providers have to keep track of the health status of fleets of complex systems. The ability to estimate the future health status of these systems and their components becomes more pivotal when seeking to efficiently operate and maintain these systems. Today, these stakeholders have access to a lot of different data sources regarding fleet, operation schedule, ambient condition, system and component information. Many different prognostic methods from different disciplines are available and will further improve henceforward. In many cases these data sources and methods function as isolated methods in their own field. This fragmentation makes a holistic prognosis very challenging in many cases. Therefore, stakeholders need information integrating methods and tools to gain an exhaustive insight into the health status development of the complex assets they are operating or maintaining, in order to make well-founded decisions regarding operation or maintenance planning. In this paper, a Python-based health index framework is presented. It enables users to integrate operation schedules of different detail levels with enriching data sources such as ambient condition data. Furthermore, it provides methods to design complex asset systems which are linked via their construction, function or degradation mechanisms/ health indices via transfer relations. It allows to monitor the asset’s condition based on operation data and to simulate different operation scenarios regarding the health index development.
{"title":"Health Index Framework for Condition Monitoring and Health Prediction","authors":"A. Kamtsiuris, F. Raddatz, Gerko Wende","doi":"10.36001/phme.2022.v7i1.3324","DOIUrl":"https://doi.org/10.36001/phme.2022.v7i1.3324","url":null,"abstract":"In the field of Maintenance, Repair and Overhaul (MRO), stakeholders such as operators or service providers have to keep track of the health status of fleets of complex systems. The ability to estimate the future health status of these systems and their components becomes more pivotal when seeking to efficiently operate and maintain these systems. Today, these stakeholders have access to a lot of different data sources regarding fleet, operation schedule, ambient condition, system and component information. Many different prognostic methods from different disciplines are available and will further improve henceforward. In many cases these data sources and methods function as isolated methods in their own field. This fragmentation makes a holistic prognosis very challenging in many cases. Therefore, stakeholders need information integrating methods and tools to gain an exhaustive insight into the health status development of the complex assets they are operating or maintaining, in order to make well-founded decisions regarding operation or maintenance planning. In this paper, a Python-based health index framework is presented. It enables users to integrate operation schedules of different detail levels with enriching data sources such as ambient condition data. Furthermore, it provides methods to design complex asset systems which are linked via their construction, function or degradation mechanisms/ health indices via transfer relations. It allows to monitor the asset’s condition based on operation data and to simulate different operation scenarios regarding the health index development.","PeriodicalId":422825,"journal":{"name":"PHM Society European Conference","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134156156","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-29DOI: 10.36001/phme.2022.v7i1.3356
Karl Lowenmark, C. Taal, Joakim Nivre, M. Liwicki, Fredrik Sandin
Annotations in condition monitoring systems contain information regarding asset history and fault characteristics in the form of unstructured text that could, if unlocked, be used for intelligent fault diagnosis. However, processing these annotations with pre-trained natural language models such as BERT is problematic due to out-of-vocabulary (OOV) technical terms, resulting in inaccurate language embeddings. Here we investigate the effect of OOV technical terms on BERT and SentenceBERT embeddings by substituting technical terms with natural language descriptions. The embeddings were computed for each annotation in a pre-processed corpus, with and without substitution. The K-Means clustering score was calculated on sentence embeddings, and a Long Short-Term Memory (LSTM) network was trained on word embeddings with the objective to recreate the output from a keywordbased annotation classifier. The K-Means score for SentenceBERT annotation embeddings improved by 40% at seven clusters by technical language substitution, and the labelling capacity of the BERT-LSTM model was improved from 88.3 to 94.2%. These results indicate that the substitution of OOV technical terms can improve the representation accuracy of the embeddings of the pre-trained BERT and SentenceBERT models, and that pre-trained language models can be used to process technical language.
{"title":"Processing of Condition Monitoring Annotations with BERT and Technical Language Substitution: A Case Study","authors":"Karl Lowenmark, C. Taal, Joakim Nivre, M. Liwicki, Fredrik Sandin","doi":"10.36001/phme.2022.v7i1.3356","DOIUrl":"https://doi.org/10.36001/phme.2022.v7i1.3356","url":null,"abstract":"Annotations in condition monitoring systems contain information regarding asset history and fault characteristics in the form of unstructured text that could, if unlocked, be used for intelligent fault diagnosis. However, processing these annotations with pre-trained natural language models such as BERT is problematic due to out-of-vocabulary (OOV) technical terms, resulting in inaccurate language embeddings. Here we investigate the effect of OOV technical terms on BERT and SentenceBERT embeddings by substituting technical terms with natural language descriptions. The embeddings were computed for each annotation in a pre-processed corpus, with and without substitution. The K-Means clustering score was calculated on sentence embeddings, and a Long Short-Term Memory (LSTM) network was trained on word embeddings with the objective to recreate the output from a keywordbased annotation classifier. The K-Means score for SentenceBERT annotation embeddings improved by 40% at seven clusters by technical language substitution, and the labelling capacity of the BERT-LSTM model was improved from 88.3 to 94.2%. These results indicate that the substitution of OOV technical terms can improve the representation accuracy of the embeddings of the pre-trained BERT and SentenceBERT models, and that pre-trained language models can be used to process technical language.","PeriodicalId":422825,"journal":{"name":"PHM Society European Conference","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123035774","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-29DOI: 10.36001/phme.2022.v7i1.3337
Ravit Ohana, R. Klein, J. Bortman
Condition based maintenance (CBM) is the preferred approach in rotating machinery and aim to replace the commonly used approach of maintenance based on service time. To achieve an effective CBM, different types of sensors should be placed in the system for condition monitoring to detect the location of the fault and its severity. In this research, a Fiber Bragg Grating (FBG) has been used for condition monitoring on spalls in deep grove ball bearings. The motivation for using these sensors is the ability to get a high-noise signal (SNR) ratio. The usage of FBG sensors is relatively new for health monitoring systems of rotating machinery. Therefore, there is not enough understanding of the strain signature measured by the FBG. To examine the phenomena in the strain signals, a physics-based model of the strain signature has been developed. In this model, two complementary models were integrated, a finite element (FE) model and a dynamic model . The strain model describes the interaction between the rolling elements (REs) and the bearing housing and simulates the strain behavior measured on the bearing housing. The simulation results are validated with strain signals measured by the FBG sensor at different stages of an endurance test. The model allows simulation of a wide range of spall lengths and describes the behavior of the strain signals for different levels of misalignment. The insights from the model enabled the development of an automatic algorithm that assess the severity of the defect and to track spall length during bearing operation, based on strain signals.
{"title":"Severity Estimation of Faulty Bearings Based on Strain Signals From Physical Models and FBG Measurements","authors":"Ravit Ohana, R. Klein, J. Bortman","doi":"10.36001/phme.2022.v7i1.3337","DOIUrl":"https://doi.org/10.36001/phme.2022.v7i1.3337","url":null,"abstract":"Condition based maintenance (CBM) is the preferred approach in rotating machinery and aim to replace the commonly used approach of maintenance based on service time. To achieve an effective CBM, different types of sensors should be placed in the system for condition monitoring to detect the location of the fault and its severity. In this research, a Fiber Bragg Grating (FBG) has been used for condition monitoring on spalls in deep grove ball bearings. The motivation for using these sensors is the ability to get a high-noise signal (SNR) ratio. The usage of FBG sensors is relatively new for health monitoring systems of rotating machinery. Therefore, there is not enough understanding of the strain signature measured by the FBG. To examine the phenomena in the strain signals, a physics-based model of the strain signature has been developed. In this model, two complementary models were integrated, a finite element (FE) model and a dynamic model . The strain model describes the interaction between the rolling elements (REs) and the bearing housing and simulates the strain behavior measured on the bearing housing. The simulation results are validated with strain signals measured by the FBG sensor at different stages of an endurance test. The model allows simulation of a wide range of spall lengths and describes the behavior of the strain signals for different levels of misalignment. The insights from the model enabled the development of an automatic algorithm that assess the severity of the defect and to track spall length during bearing operation, based on strain signals.","PeriodicalId":422825,"journal":{"name":"PHM Society European Conference","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128503785","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-29DOI: 10.36001/phme.2022.v7i1.3329
Maximilian-Peter Radtke, Jürgen Bock
In the recent past deep learning approaches have achieved some remarkable results in the area of fault diagnostics and anomaly detection. Nevertheless, these algorithms rely on large amounts of data, which is often not available, and produce outputs, which are hard to interpret. These deficiencies make real life applications difficult. Before the broad success of deep learning machine faults were often classified using domain expert knowledge based on experience and physical models. In comparison, these approaches only require small amounts of data and produce highly interpretable results. On the downside, however, they struggle to predict unexpected patterns hidden in data. Merging these two concepts promises to increase accuracy, robustness and interpretability of models. In this paper we present a hybrid approach to combine expert knowledge with deep learning and evaluate it on rolling element bearing fault detection. First, we create a knowledge base for fault classification derived from the expected physical attributes of different faults in the envelope spectrum of vibration signals. This knowledge is used to derive a similarity function for comparing input signals to expected faulty signals. Afterwards, the similarity measure is incorporated into different neural networks using a Logic Tensor Network (LTN). This enables logical reasoning in the loss function, in which we aim to mimic the decision process of an expert analyzing the input data. Further, we extend LTNs by weight schedules for axiom groups. We show that our approach outperforms the baseline models on two bearing fault data sets with different attributes and directly gives a better understanding of whether or not fault signals are influenced by other effects or behave as expected.
{"title":"Expert Knowledge Induced Logic Tensor Networks: A Bearing Fault Diagnosis Case Study","authors":"Maximilian-Peter Radtke, Jürgen Bock","doi":"10.36001/phme.2022.v7i1.3329","DOIUrl":"https://doi.org/10.36001/phme.2022.v7i1.3329","url":null,"abstract":"In the recent past deep learning approaches have achieved some remarkable results in the area of fault diagnostics and anomaly detection. Nevertheless, these algorithms rely on large amounts of data, which is often not available, and produce outputs, which are hard to interpret. These deficiencies make real life applications difficult. Before the broad success of deep learning machine faults were often classified using domain expert knowledge based on experience and physical models. In comparison, these approaches only require small amounts of data and produce highly interpretable results. On the downside, however, they struggle to predict unexpected patterns hidden in data. Merging these two concepts promises to increase accuracy, robustness and interpretability of models. In this paper we present a hybrid approach to combine expert knowledge with deep learning and evaluate it on rolling element bearing fault detection. First, we create a knowledge base for fault classification derived from the expected physical attributes of different faults in the envelope spectrum of vibration signals. This knowledge is used to derive a similarity function for comparing input signals to expected faulty signals. Afterwards, the similarity measure is incorporated into different neural networks using a Logic Tensor Network (LTN). This enables logical reasoning in the loss function, in which we aim to mimic the decision process of an expert analyzing the input data. Further, we extend LTNs by weight schedules for axiom groups. We show that our approach outperforms the baseline models on two bearing fault data sets with different attributes and directly gives a better understanding of whether or not fault signals are influenced by other effects or behave as expected.","PeriodicalId":422825,"journal":{"name":"PHM Society European Conference","volume":"101 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124929509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-29DOI: 10.36001/phme.2022.v7i1.3360
Pablo Del Moral, Sławomir Nowaczyk, Sepideh Pashami
One of the main challenges for predictive maintenance in real applications is the quality of the data, especially the labels. In this paper, we propose a methodology to filter out the misleading labels that harm the performance of Machine Learning models. Ideally, predictive maintenance would be based on the information of when a fault has occurred in a machine and what specific type of fault it was. Then, we could train machine learning models to identify the symptoms of such fault before it leads to a breakdown. However, in many industrial applications, this information is not available. Instead, we approximate it using a log of component replacements, usually coming from the sales or maintenance departments. The repair history provides reliable labels for fault prediction models only if the replaced component was indeed faulty, with symptoms captured by collected data, and it was going to lead to a breakdown. However, very often, at least for complex equipment, this assumption does not hold. Models trained using unreliable labels will then, necessarily, fail. We demonstrate that filtering misleading labels leads to improved results. Our central claim is that the same fault, happening several times, should have similar symptoms in the data; thus, we can train a model to predict them. On the contrary, replacements of the same component that do not exhibit similar symptoms will be confusing and harm the ML models. Therefore, we aim to filter the maintenance operations, keeping only those that can be used to predict each other. Suppose we can train a successful model using the data before a component replacement to predict another component replacement. In that case, those maintenance operations must be motivated by the same, or a very similar, type of fault. We test this approach on a real scenario using data from a fleet of sterilizers deployed in hospitals. The data includes sensor readings from the machines describing their operations and the service logs indicating the replacement of components when the manufacturing company performs the service. Since sterilizers are complex machines consisting of many components and systems interacting with each other, there is the possibility of faults happening simultaneously.
{"title":"Filtering Misleading Repair Log Labels to Improve Predictive Maintenance Models","authors":"Pablo Del Moral, Sławomir Nowaczyk, Sepideh Pashami","doi":"10.36001/phme.2022.v7i1.3360","DOIUrl":"https://doi.org/10.36001/phme.2022.v7i1.3360","url":null,"abstract":"One of the main challenges for predictive maintenance in real applications is the quality of the data, especially the labels. In this paper, we propose a methodology to filter out the misleading labels that harm the performance of Machine Learning models. Ideally, predictive maintenance would be based on the information of when a fault has occurred in a machine and what specific type of fault it was. Then, we could train machine learning models to identify the symptoms of such fault before it leads to a breakdown. However, in many industrial applications, this information is not available. Instead, we approximate it using a log of component replacements, usually coming from the sales or maintenance departments. The repair history provides reliable labels for fault prediction models only if the replaced component was indeed faulty, with symptoms captured by collected data, and it was going to lead to a breakdown.\u0000However, very often, at least for complex equipment, this assumption does not hold. Models trained using unreliable labels will then, necessarily, fail. We demonstrate that filtering misleading labels leads to improved results. Our central claim is that the same fault, happening several times, should have similar symptoms in the data; thus, we can train a model to predict them. On the contrary, replacements of the same component that do not exhibit similar symptoms will be confusing and harm the ML models. Therefore, we aim to filter the maintenance operations, keeping only those that can be used to predict each other. Suppose we can train a successful model using the data before a component replacement to predict another component replacement. In that case, those maintenance operations must be motivated by the same, or a very similar, type of fault.\u0000We test this approach on a real scenario using data from a fleet of sterilizers deployed in hospitals. The data includes sensor readings from the machines describing their operations and the service logs indicating the replacement of components when the manufacturing company performs the service. Since sterilizers are complex machines consisting of many components and systems interacting with each other, there is the possibility of faults happening simultaneously.","PeriodicalId":422825,"journal":{"name":"PHM Society European Conference","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130878253","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-29DOI: 10.36001/phme.2022.v7i1.3312
Friedrich von Bülow, Tobias Meisen
Due to the global transition to electromobility and the associated increased use of high-performance batteries, research is increasingly focused on estimating and forecasting the state of health (SOH) of lithium-ion batteries. Several data-intensive and well-performing methods for SOH forecasting have been introduced. However, these approaches are only reliable for new battery types, e.g., with a new cell chemistry, if a sufficient amount of training data is given, which is rarely the case. A promising approach is to transfer an established model of another battery type to the new battery type, using only a small amount of data of the new battery type. Such methods in machine learning are known as transfer learning. The usefulness and applicability of transfer learning and its underlying methods have been very successfully demonstrated in various fields, such as computer vision and natural language processing. Heterogeneity in battery systems, such as differences in rated capacity, cell cathode materials, as well as applied stress from use, necessitate transfer learning concepts for data-based battery SOH forecasting models. Hereby, the general electrochemical behavior of lithium-ion batteries, as a major common characteristic, supposedly provides an excellent starting point for a transfer learning approach for SOH forecasting models. In this paper, we present a transfer learning approach for SOH forecasting models using a multilayer perceptron (MLP). We apply and evaluate it on the method presented by von Bülow, Mentz, and Meisen (2021) using five battery datasets. In this regard, we investigate the optimal conditions and settings for the development of transfer learning with respect to suitable data from the target domain, as well as hyperparameters such as learning rate and frozen layers. We show that for the transfer of a SOH forecasting model to a new battery type it is more beneficial to have data of few old batteries, compared to data of many new batteries, especially in the case of superlinear degradation with knee points. Contrarily to computer vision freezing no layers is preferable in 95% of the experimental scenarios.
由于全球向电动汽车的过渡以及高性能电池的使用增加，研究越来越关注于锂离子电池的健康状态(SOH)的估计和预测。介绍了几种数据密集且性能良好的SOH预测方法。然而，这些方法仅适用于新电池类型，例如，如果提供了足够数量的训练数据，则具有新的电池化学性质，而这种情况很少发生。一种很有前景的方法是将另一种电池类型的已建立模型转移到新电池类型，只使用少量新电池类型的数据。这种机器学习方法被称为迁移学习。迁移学习及其基础方法的有用性和适用性已经在计算机视觉和自然语言处理等各个领域得到了非常成功的证明。电池系统的异质性，如额定容量、电池正极材料以及使用过程中的应用应力的差异，需要基于数据的电池SOH预测模型的迁移学习概念。因此，锂离子电池的一般电化学行为作为一个主要的共同特征，可以为SOH预测模型的迁移学习方法提供一个很好的起点。本文提出了一种基于多层感知器(MLP)的SOH预测模型迁移学习方法。我们在von b low, Mentz和Meisen(2021)使用五个电池数据集提出的方法上应用并评估它。在这方面，我们研究了迁移学习发展的最佳条件和设置，涉及目标域的合适数据，以及学习率和冻结层等超参数。我们表明，对于将SOH预测模型转移到新电池类型而言，拥有少量旧电池的数据比拥有许多新电池的数据更有利，特别是在具有膝盖点的超线性退化的情况下。与计算机视觉相反，在95%的实验场景中，冻结无层是更可取的。
{"title":"State of Health Forecasting of Heterogeneous Lithium-ion Battery Types and Operation Enabled by Transfer Learning","authors":"Friedrich von Bülow, Tobias Meisen","doi":"10.36001/phme.2022.v7i1.3312","DOIUrl":"https://doi.org/10.36001/phme.2022.v7i1.3312","url":null,"abstract":"Due to the global transition to electromobility and the associated increased use of high-performance batteries, research is increasingly focused on estimating and forecasting the state of health (SOH) of lithium-ion batteries. Several data-intensive and well-performing methods for SOH forecasting have been introduced. However, these approaches are only reliable for new battery types, e.g., with a new cell chemistry, if a sufficient amount of training data is given, which is rarely the case. A promising approach is to transfer an established model of another battery type to the new battery type, using only a small amount of data of the new battery type. Such methods in machine learning are known as transfer learning. The usefulness and applicability of transfer learning and its underlying methods have been very successfully demonstrated in various fields, such as computer vision and natural language processing. Heterogeneity in battery systems, such as differences in rated capacity, cell cathode materials, as well as applied stress from use, necessitate transfer learning concepts for data-based battery SOH forecasting models. Hereby, the general electrochemical behavior of lithium-ion batteries, as a major common characteristic, supposedly provides an excellent starting point for a transfer learning approach for SOH forecasting models. In this paper, we present a transfer learning approach for SOH forecasting models using a multilayer perceptron (MLP). We apply and evaluate it on the method presented by von Bülow, Mentz, and Meisen (2021) using five battery datasets. In this regard, we investigate the optimal conditions and settings for the development of transfer learning with respect to suitable data from the target domain, as well as hyperparameters such as learning rate and frozen layers. We show that for the transfer of a SOH forecasting model to a new battery type it is more beneficial to have data of few old batteries, compared to data of many new batteries, especially in the case of superlinear degradation with knee points. Contrarily to computer vision freezing no layers is preferable in 95% of the experimental scenarios.","PeriodicalId":422825,"journal":{"name":"PHM Society European Conference","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130698259","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-29DOI: 10.36001/phme.2022.v7i1.3296
Renato Murata, Louis Thioulouse, J. Marzat, H. Piet-Lahanier, M. Galeotta, Rancois Farago
A possible way to increase the reliability and availability of a system is to apply an Active Fault Tolerant Control (AFTC) algorithm. This thesis aims to use this algorithm in a multiengine propulsive cluster with sensor and actuator faults. First, a Health Monitoring System (HMS) will be developed to monitor the entire propulsive cluster. The HMS will use model-based fault diagnosis techniques. Then, in case of actuator faults, the cluster will be reconfigured to minimize its effects. The reconfiguration can be made by using control allocation or modifying the control law of the engine. A simulation model of the entire cluster is under development. The model simulates the whole system, including the propellant feeding system, engines, and mechanical system. It will be used to study the effect of different faults on the system and compare different reconfiguration strategies.
{"title":"Diagnosis and Fault-Tolerant Control for a Multi-Engine Cluster of a Reusable Launcher with Sensor and Actuator Faults","authors":"Renato Murata, Louis Thioulouse, J. Marzat, H. Piet-Lahanier, M. Galeotta, Rancois Farago","doi":"10.36001/phme.2022.v7i1.3296","DOIUrl":"https://doi.org/10.36001/phme.2022.v7i1.3296","url":null,"abstract":"A possible way to increase the reliability and availability of a system is to apply an Active Fault Tolerant Control (AFTC) algorithm. This thesis aims to use this algorithm in a multiengine propulsive cluster with sensor and actuator faults. First, a Health Monitoring System (HMS) will be developed to monitor the entire propulsive cluster. The HMS will use model-based fault diagnosis techniques. Then, in case of actuator faults, the cluster will be reconfigured to minimize its effects. The reconfiguration can be made by using control allocation or modifying the control law of the engine. A simulation model of the entire cluster is under development. The model simulates the whole system, including the propellant feeding system, engines, and mechanical system. It will be used to study the effect of different faults on the system and compare different reconfiguration strategies.","PeriodicalId":422825,"journal":{"name":"PHM Society European Conference","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122893568","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-06-29DOI: 10.36001/phme.2022.v7i1.3362
Karolina Sobczak-Oramus, A. Mosallam, Caner Basci, Jinlong Kang
Logging tools widely used in the oil and gas industry are exposed to demanding environmental conditions that can lead to faster degradation and unexpected failures. These events can reduce productivity, delay deliverables, or even bring entire drilling operations to an end. However, such accidents can be avoided using a prognostics and health management approach. This paper presents a data-driven fault detection method for transmitter in logging-while-drilling tool adopting a support vector machine classifier. The health analyzer determines the component’s physical condition in just a few minutes, demonstrating an exceptional value for both field and maintenance engineers. This work is part of a long-term project aimed at constructing a digital fleet management system for downhole testing tools.
{"title":"Data-Driven Fault Detection for Transmitter in Logging-While-Drilling Tool","authors":"Karolina Sobczak-Oramus, A. Mosallam, Caner Basci, Jinlong Kang","doi":"10.36001/phme.2022.v7i1.3362","DOIUrl":"https://doi.org/10.36001/phme.2022.v7i1.3362","url":null,"abstract":"Logging tools widely used in the oil and gas industry are exposed to demanding environmental conditions that can lead to faster degradation and unexpected failures. These events can reduce productivity, delay deliverables, or even bring entire drilling operations to an end. However, such accidents can be avoided using a prognostics and health management approach. This paper presents a data-driven fault detection method for transmitter in logging-while-drilling tool adopting a support vector machine classifier. The health analyzer determines the component’s physical condition in just a few minutes, demonstrating an exceptional value for both field and maintenance engineers. This work is part of a long-term project aimed at constructing a digital fleet management system for downhole testing tools.","PeriodicalId":422825,"journal":{"name":"PHM Society European Conference","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121857299","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：