{"title":"Recurrent neural networks for remaining useful life estimation","authors":"F. Heimes","doi":"10.1109/PHM.2008.4711422","DOIUrl":"https://doi.org/10.1109/PHM.2008.4711422","url":null,"abstract":"This paper presents an approach and solution to the IEEE 2008 Prognostics and Health Management conference challenge problem. The solution utilizes an advanced recurrent neural network architecture to estimate the remaining useful life of the system. The recurrent neural network is trained with back-propagation through time gradient calculations, an Extended Kalman Filter training method, and evolutionary algorithms to generate an accurate and compact algorithm. This solution placed second overall in the competition with a very small margin between the first and second place finishers.","PeriodicalId":377635,"journal":{"name":"2008 International Conference on Prognostics and Health Management","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123503023","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}
{"title":"Accommodation of partial actuator faults using output feedback based adaptive robust control","authors":"S. Gayaka, B. Yao","doi":"10.1109/PHM.2008.4711449","DOIUrl":"https://doi.org/10.1109/PHM.2008.4711449","url":null,"abstract":"This work considers the problem of partial actuator fault accommodation. The faulty actuators are characterized by a loss in efficiency, which can potentially lead to system instability or degraded system performance. Conventional model reference adaptive control has been used to solve this problem in the literature, which provides a natural setting to solve such problems by virtue of its online learning capabilities. Unfortunately, such techniques can lead to unbounded states when output disturbances and other modeling uncertainties are considered. In the present work, we make the problem more practical by explicitly taking into account these factors. Furthermore, we assume that faults occur at unknown instants of time and only limited state information is available. For solving this problem, we develop an output feedback adaptive robust control (ARC) based technique, which can effectively deal with all the constraints and uncertainties present in the system. The technique combines the output feedback based adaptive backstepping with robust control techniques and uses discontinuous projection to ensure the unknown parameter estimates stay within a known convex set. Comparative simulation studies are performed on a linearized Boeing 747 model to evaluate performance of the proposed scheme versus conventional MRAC based technique.","PeriodicalId":377635,"journal":{"name":"2008 International Conference on Prognostics and Health Management","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121596680","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}
{"title":"A statistical model for estimating probability of crack detection","authors":"A. Coppe, R. Haftka, Nam-Ho Kim, C. Bès","doi":"10.1109/PHM.2008.4711418","DOIUrl":"https://doi.org/10.1109/PHM.2008.4711418","url":null,"abstract":"In the inspection of aircraft structures, the probability of detection has typically been determined based on the size of damage alone. However, the inspection process involves randomness due to variability in inspection conditions (including inspectorpsilas competence), as well as difficulties associated with the location and type of damages. To demonstrate the effects of these other factors, we present a simple model from the assumption that for each combination of crack location and inspector there is a threshold crack size such that all cracks above this size will be detected and all below that size will be missed. The proposed model adjusts the threshold crack size according to the difficulty associated with the crack location and the competence of inspectors. The model is then used to fit 2,603 detection events reported for 43 panels inspected by 62 technicians in an Air Force study. The threshold increments by location and inspector are obtained by maximizing the matching percentage in detection events between the model and the experiment. We first use 62 inspector thresholds only and find the best matching percentage of 78%. It is further increased to 81% when both inspector and location thresholds are considered. For comparison, the matching percentage using crack size alone is only 55%. We then add randomness to the process in order to include inconsistency on the part of inspectors. Replicating the observed inconsistency reduces the matching to about 72%. We conclude that most of the randomness in manual inspections is due to the circumstances of the inspections. We speculate that much of this randomness will be eliminated by automated structural health monitoring (SHM), which will be an important benefit of SHM.","PeriodicalId":377635,"journal":{"name":"2008 International Conference on Prognostics and Health Management","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115911505","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}
Bin Zhang, Chris Sconyers, C. Byington, R. Patrick, Marcos E. Orchard, George Vachtsevanos
{"title":"Anomaly detection: A robust approach to detection of unanticipated faults","authors":"Bin Zhang, Chris Sconyers, C. Byington, R. Patrick, Marcos E. Orchard, George Vachtsevanos","doi":"10.1109/PHM.2008.4711445","DOIUrl":"https://doi.org/10.1109/PHM.2008.4711445","url":null,"abstract":"This paper introduces a methodology to detect as early as possible with specified degree of confidence and prescribed false alarm rate an anomaly or novelty (incipient failure) associated with critical components/subsystems of an engineered system that is configured to monitor continuously its health status. Innovative features of the enabling technologies include a Bayesian estimation framework, called particle filtering, that employs features or condition indicators derived from sensor data in combination with simple models of the systempsilas degrading state to detect a deviation or discrepancy between a baseline (no-fault) distribution and its current counterpart. The scheme provides the probability of abnormal condition and the probability of false alarm. The presence of an anomaly is confirmed for a given confidence level. The efficacy of the proposed anomaly detection architecture is illustrated with test data acquired from components typically found on aircraft and monitored via a test rig appropriately instrumented.","PeriodicalId":377635,"journal":{"name":"2008 International Conference on Prognostics and Health Management","volume":"123 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117318572","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}
{"title":"A framework for systematic benchmarking of monitoring and diagnostic systems","authors":"T. Kurtoglu, O. Mengshoel, S. Poll","doi":"10.1109/PHM.2008.4711454","DOIUrl":"https://doi.org/10.1109/PHM.2008.4711454","url":null,"abstract":"In this paper, we present an architecture and a formal framework to be used for systematic benchmarking of monitoring and diagnostic systems and for producing comparable performance assessments of different diagnostic technologies. The framework defines a number of standardized specifications, which include a fault catalog, a library of modular test scenarios, and a common protocol for gathering and processing diagnostic data. At the center of the framework are 13 benchmarking metric definitions. The calculation of metrics is illustrated on a probabilistic model-based diagnosis algorithm utilizing Bayesian reasoning techniques. The diagnosed system is a real-world electrical power system, namely the Advanced Diagnostics and Prognostics Testbed (ADAPT) developed and located at the NASA Ames Research Center. The proposed benchmarking approach shows how to generate realistic diagnostic data sets for large-scale, complex engineering systems, and how the generated experimental data can be used to enable ldquoapples to applesrdquo assessments of the effectiveness of different diagnostic and monitoring algorithms.","PeriodicalId":377635,"journal":{"name":"2008 International Conference on Prognostics and Health Management","volume":"112 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127105623","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}
{"title":"Predicting the Time To Failure of a randomly degrading component by a hybrid Monte Carlo and possibilistic method","authors":"P. Baraldi, I. Popescu, Enrico Zio","doi":"10.1109/PHM.2008.4711455","DOIUrl":"https://doi.org/10.1109/PHM.2008.4711455","url":null,"abstract":"The main goal of a prognostic system is to estimate the Time To Failure (TTF) of a structure, system or component (SSC), i.e. the lifetime remaining between the present and the instance when it can no longer perform its function. Such prediction on the time of loss of functionality is typically done on the basis of measurements of parameters representative of the SSC condition state and degradation process. Uncertainties from two different sources affect the prediction: randomness due to inherent variability in the SSC degradation behavior (aleatory uncertainty) and imprecision due to incomplete knowledge and information on the SSC stress and strength characteristics (epistemic uncertainty). Such uncertainties must be adequately represented and propagated in order for the prognostic results to have operative significance, e.g. in terms of maintenance and renovation decisions. This work illustrates an hybrid Monte Carlo and possibilistic method for the representation and propagation of aleatory and epistemic uncertainties, with reference to a case study of a component which is randomly degrading in time according to a stochastic fatigue crack growth model of literature. The crack depth is assumed as representative parameter of the component condition state and its limit value beyond which failure occurs is assumed to be affected by epistemic uncertainty, as represented by a possibility distribution.","PeriodicalId":377635,"journal":{"name":"2008 International Conference on Prognostics and Health Management","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122632448","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}
F. Peysson, M. Ouladsine, R. Outbib, J. Léger, O. Myx, C. Allemand
{"title":"Damage trajectory analysis based prognostic","authors":"F. Peysson, M. Ouladsine, R. Outbib, J. Léger, O. Myx, C. Allemand","doi":"10.1109/PHM.2008.4711428","DOIUrl":"https://doi.org/10.1109/PHM.2008.4711428","url":null,"abstract":"To obtain high availability with reduced life cycle total ownership costs, classical maintenance policies are not sufficient. Indeed these polices do not allow us to maintain just when its necessary because they are not available to plan the current system damage state in the future. To predict accurately and precisely the future system damage state, it is necessary to take into account how and where the system is used in order to analyze the damage trajectory. The paper presents a methodology based on the system decomposition in three levels: environment, mission and process, to predict the future damage state by tracking its various damage trajectories and thus to know whether the system is able to accomplish its mission in time by using system current damage state and its future use.","PeriodicalId":377635,"journal":{"name":"2008 International Conference on Prognostics and Health Management","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131515052","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}
{"title":"A similarity-based prognostics approach for Remaining Useful Life estimation of engineered systems","authors":"Tianyi Wang, Jianbo Yu, David Siegel, J. Lee","doi":"10.1109/PHM.2008.4711421","DOIUrl":"https://doi.org/10.1109/PHM.2008.4711421","url":null,"abstract":"This paper presents a similarity-based approach for estimating the Remaining Useful Life (RUL) in prognostics. The approach is especially suitable for situations in which abundant run-to-failure data for an engineered system are available. Data from multiple units of the same system are used to create a library of degradation patterns. When estimating the RUL of a test unit, the data from it will be matched to those patterns in the library and the actual life of those matched units will be used as the basis of estimation. This approach is used to tackle the data challenge problem defined by the 2008 PHM Data Challenge Competition, in which, run-to-failure data of an unspecified engineered system are provided and the RUL of a set of test units will be estimated. Results show that the similarity-based approach is very effective in performing RUL estimation.","PeriodicalId":377635,"journal":{"name":"2008 International Conference on Prognostics and Health Management","volume":"88 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132506889","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}
{"title":"Improving preciseness of time to failure predictions: Application to APU starter","authors":"S. Létourneau, Chunsheng Yang, Zhenkai Liu","doi":"10.1109/PHM.2008.4711453","DOIUrl":"https://doi.org/10.1109/PHM.2008.4711453","url":null,"abstract":"Despite the availability of huge amounts of data and a variety of powerful data analysis methods, prognostic models are still often failing to provide accurate and precise time to failure estimations. This paper addresses this problem by integrating several machine learning algorithms. The approach proposed relies on a classification system to determine the likelihood of component failures and to provide rough indications of remaining life. It then introduces clustering and SVM-based local regression to refine the time to failure estimations provided by the classification system. The paper illustrates the applicability of the proposed approach through a real world aerospace application and discusses data pre-processing requirements. The preliminary results show that the proposed method can reduce uncertainty in time to failure estimates, which in turn helps augment the usefulness of prognostics.","PeriodicalId":377635,"journal":{"name":"2008 International Conference on Prognostics and Health Management","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117351142","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}
D.W. Brown, D. L. Edwards, G. Georgoulas, Bo Zhang, G. Vachtsevanos
{"title":"Real-time fault detection and accommodation for COTS resolver position sensors","authors":"D.W. Brown, D. L. Edwards, G. Georgoulas, Bo Zhang, G. Vachtsevanos","doi":"10.1109/PHM.2008.4711450","DOIUrl":"https://doi.org/10.1109/PHM.2008.4711450","url":null,"abstract":"Resolver sensors are utilized as absolute position transducers to control the position and speed of actuators in many flight critical applications where robustness, accuracy and ability to operate in extreme environmental conditions are required. To ensure these requirements, several designs for self-diagnosing sensors were proposed in the past. However, such designs require additional instrumentation leading to further development and certification costs. To mitigate such costs, a methodology was sought to achieve similar self-diagnosis capabilities that can be retrofitted to already existing commercial off-the-shelf (COTS) resolver sensors. To achieve this goal, this paper proposes a new approach for real-time tracking of resolver faults with the ability to perform real-time fault detection. However, the additional benefit of this approach is the ability to accommodate faults in real-time once a fault is detected. The primary fault studied in this paper is resolver channel mismatch. A formulation based on the physical operating principles for a resolver sensor is provided. Diagnostic measures are identified for use in statistical-based fault-detection routines. Then, estimates for the resolver mismatch are tracked using a time-varying Kalman filter. Fault accommodation is achieved by applying the tracked estimates to adjust for the resolver mismatch. Finally, the fault detection and accommodation routines are evaluated in Simulink for an electro-mechanical actuator (EMA).","PeriodicalId":377635,"journal":{"name":"2008 International Conference on Prognostics and Health Management","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128129046","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}