Change search
Refine search result
1 - 31 of 31
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Rows per page
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sort
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
Select
The maximal number of hits you can export is 250. When you want to export more records please use the Create feeds function.
  • 1.
    Arias Chao, Manuel
    et al.
    Chair of Intelligent Maintenance Systems, ETH Zürich, 8093 Zürich, Switzerland.
    Kulkarni, Chetan
    KBR, Inc., NASA Ames Research Center, Mountain View, CA 94035, USA.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. PARC, Intelligent Systems Lab, Palo Alto, CA 94043, USA.
    Fink, Olga
    Chair of Intelligent Maintenance Systems, ETH Zürich, 8093 Zürich, Switzerland.
    Aircraft Engine Run-to-Failure Dataset under Real Flight Conditions for Prognostics and Diagnostics2021In: Data, E-ISSN 2306-5729, Vol. 6, no 1, p. 5-5Article in journal (Refereed)
    Abstract [en]

    A key enabler of intelligent maintenance systems is the ability to predict the remaining useful lifetime (RUL) of its components, i.e., prognostics. The development of data-driven prognostics models requires datasets with run-to-failure trajectories. However, large representative run-to-failure datasets are often unavailable in real applications because failures are rare in many safety-critical systems. To foster the development of prognostics methods, we develop a new realistic dataset of run-to-failure trajectories for a fleet of aircraft engines under real flight conditions. The dataset was generated with the Commercial Modular Aero-Propulsion System Simulation (CMAPSS) model developed at NASA. The damage propagation modelling used in this dataset builds on the modelling strategy from previous work and incorporates two new levels of fidelity. First, it considers real flight conditions as recorded on board of a commercial jet. Second, it extends the degradation modelling by relating the degradation process to its operation history. This dataset also provides the health, respectively, fault class. Therefore, besides its applicability to prognostics problems, the dataset can be used for fault diagnostics. 

  • 2.
    Arias Chao, Manuel
    et al.
    Chair of Intelligent Maintenance Systems, ETH Zurich, Switzerland.
    Kulkarni, Chetan
    KBR, Inc., NASA Ames Research Center, USA.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics.
    Fink, Olga
    Chair of Intelligent Maintenance Systems, ETH Zurich, Switzerland.
    Fusing physics-based and deep learning models for prognostics2022In: Reliability Engineering & System Safety, ISSN 0951-8320, E-ISSN 1879-0836, Vol. 217, article id 107961Article in journal (Refereed)
    Abstract [en]

    Physics-based and data-driven models for remaining useful lifetime (RUL) prediction typically suffer from two major challenges that limit their applicability to complex real-world domains: (1) the incompleteness of physics-based models and (2) the limited representativeness of the training dataset for data-driven models. Combining the advantages of these two approaches while overcoming some of their limitations, we propose a novel hybrid framework for fusing the information from physics-based performance models with deep learning algorithms for prognostics of complex safety-critical systems. In the proposed framework, we use physics-based performance models to infer unobservable model parameters related to a system’s components health by solving a calibration problem. These parameters are subsequently combined with sensor readings and used as input to a deep neural network, thereby generating a data-driven prognostics model with physics-augmented features. The performance of the hybrid framework is evaluated on an extensive case study comprising run-to-failure degradation trajectories from a fleet of nine turbofan engines under real flight conditions. The experimental results show that the hybrid framework outperforms purely data-driven approaches by extending the prediction horizon by nearly 127%. Furthermore, it requires less training data and is less sensitive to the limited representativeness of the dataset as compared to purely data-driven approaches. Furthermore, we demonstrated the feasibility of the proposed framework on the original CMAPSS dataset, thereby confirming its superior performance.

  • 3.
    Baptista, Marcia L.
    et al.
    Delft University of Technology (TU Delft), Mekelweg 5, 2628 CD Delft, the Netherlands.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center (PARC), Palo Alto CA 94304, USA.
    Henriques, Elsa M.P.
    University of Lisbon - Instituto Superior Tecnico (IST), Av. Rovisco Pais nº1, 1049-001 Lisbon, Portugal.
    Relation between prognostics predictor evaluation metrics and local interpretability SHAP values2022In: Artificial Intelligence, ISSN 0004-3702, E-ISSN 1872-7921, Vol. 306, article id 103667Article in journal (Refereed)
    Abstract [en]

    Maintenance decisions in domains such as aeronautics are becoming increasingly dependent on being able to predict the failure of components and systems. When data-driven techniques are used for this prognostic task, they often face headwinds due to their perceived lack of interpretability. To address this issue, this paper examines how features used in a data-driven prognostic approach correlate with established metrics of monotonicity, trendability, and prognosability. In particular, we use the SHAP model (SHapley Additive exPlanations) from the field of eXplainable Artificial Intelligence (XAI) to analyze the outcome of three increasingly complex algorithms: Linear Regression, Multi-Layer Perceptron, and Echo State Network. Our goal is to test the hypothesis that the prognostics metrics correlate with the SHAP model's explanations, i.e., the SHAP values. We use baseline data from a standard data set that contains several hundred run-to-failure trajectories for jet engines. The results indicate that SHAP values track very closely with these metrics with differences observed between the models that support the assertion that model complexity is a significant factor to consider when explainability is a consideration in prognostics.

  • 4.
    Baptista, Marcia L.
    et al.
    IDMEC, Instituto Superior Tecnico, University of Lisbon, Av. Rovisco Pais 1, 1049-001 Lisboa, Portugal.
    Henriques, Elsa M.P.
    IDMEC, Instituto Superior Tecnico, University of Lisbon, Av. Rovisco Pais 1, 1049-001 Lisboa, Portugal.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. PARC, Palo Alto, CA 94304, United States.
    More effective prognostics with elbow point detection and deep learning2021In: Mechanical systems and signal processing, ISSN 0888-3270, E-ISSN 1096-1216, Vol. 146, article id 106987Article in journal (Refereed)
    Abstract [en]

    Prior to failure, most systems exhibit signs of changed characteristics. The early detection of this change is important to remaining useful life estimation. To have the ability to detect the inflection point or “elbow point” of an asset, i.e. the point of the degradation curve that marks the transition from nominal to faulty condition, can enable more sophisticated prognostics because this divide and conquer tactic allows the prediction to focus on the window before failure when significant changes are being expected. In this work, we compare prognostics with and without change point detection. We use different recurrent neural network techniques (standard recurrent neural network, long short-term memory and gated recurrent unit) to find the elbow point location. The actual estimation of the remaining time to failure is based on the echo state network, a state-of-the-art approach in prognostics. Two different experiments are performed on simulated data obtained from NASA Ames prognostics repository. We first compare the performance of the elbow point detectors based on recurrent neural networks against three baseline models: the Z-test, multi-layer perceptron and random forests. Results indicate that recurrent neural networks can outperform the baseline approaches. In the second experiment, the best elbow detection model, the gated recurrent unit, is integrated within an echo state network, with a significant increase in overall performance in terms of remaining useful life estimation.

  • 5.
    Baptista, Marcia Lourenco
    et al.
    Air Transport and Operations, Faculty of Aerospace Engineering, Delft University of Technology, Delft, Netherlands.
    Henriques, Elsa M.P.
    LAETA, IDMEC, Instituto Superior Tecnico, University of Lisbon, Av. Rovisco Pais 1, 1049-001 Lisboa, Portugal.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center, Palo Alto, CA, 94304, United States.
    A self-organizing map and a normalizing multi-layer perceptron approach to baselining in prognostics under dynamic regimes2021In: Neurocomputing, ISSN 0925-2312, E-ISSN 1872-8286, Vol. 456, p. 268-287Article in journal (Refereed)
    Abstract [en]

    When the influence of changing operational and environmental conditions is not factored out, it can be dificult to observe a clear deterioration path. This can significantly affect the task of prognostics and other analytic operations. To address this issue, it is necessary to baseline the data, typically by first finding the operating regimes and then normalizing the data within each regime. In this paper, we propose the use of machine learning techniques to perform baselining. A self-organizing map identifies the regimes, and a multi-layer perceptron normalizes the data based on the detected regimes. Tests are performed on the C-MAPSS data. The approach is capable of producing similar results to classical methods without the need to specify in advance the number of regimes and the explicit computation of the statistical properties of a hold-out dataset. Importantly, the techniques can be integrated into a deep learning system to perform prognostics in a single pass.

  • 6.
    Bektas, Oguz
    et al.
    Warwick Manufacturing Group, University of Warwick, Coventry, UK.
    Jones, Jeffrey A.
    Warwick Manufacturing Group, University of Warwick, Coventry, UK.
    Sankararaman, Shankar
    Data Science and Analytics Manager,Pricewaterhouse Cooper, San Jose, USA.
    Roychoudhury, Indranil
    Stinger Ghaffarian Technologies, Inc.NASA Ames Research Center, Moffett Field, USA.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. NASA Ames Research Center, Moffett Field, USA.
    A neural network filtering approach for similarity-based remaining useful life estimation2019In: The International Journal of Advanced Manufacturing Technology, ISSN 0268-3768, E-ISSN 1433-3015, Vol. 101, no 1-4, p. 87-103Article in journal (Refereed)
    Abstract [en]

    The role of prognostics and health management is ever more prevalent with advanced techniques of estimation methods. However, data processing and remaining useful life prediction algorithms are often very different. Some difficulties in accurate prediction can be tackled by redefining raw data parameters into more meaningful and comprehensive health level indicators that will then provide performance information. Proper data processing has a significant importance on remaining useful life predictions, for example, to deal with data limitations or/and multi-regime operating conditions. The framework proposed in this paper considers a similarity-based prognostic algorithm that is fed by the use of data normalisation and filtering methods for operational trajectories of complex systems. This is combined with a data-driven prognostic technique based on feed-forward neural networks with multi-regime normalisation. In particular, the paper takes a close look at how pre-processing methods affect algorithm performance. The work presented herein shows a conceptual prognostic framework that overcomes challenges presented by short-term test datasets and that increases the prediction performance with regards to prognostic metrics.

    Download full text (pdf)
    fulltext
  • 7.
    Bektas, Oguz
    et al.
    Warwick Manufacturing Group, University of Warwick, Coventry, United Kingdom.
    Jones, Jeffrey A.
    Warwick Manufacturing Group, University of Warwick, Coventry, United Kingdom.
    Sankararaman, Shankar
    Pricewaterhouse Cooper, San Jose, CA, United States.
    Roychoudhury, Indranil
    Stinger Ghaffarian Technologies, Inc., NASA Ames Research Center, Moffett Field, CA, United States.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. NASA Ames Research Center, Moffett Field, CA, United States.
    A neural network framework for similarity-based prognostics2019In: MethodsX, ISSN 1258-780X, E-ISSN 2215-0161, Vol. 6, p. 383-390Article in journal (Refereed)
    Abstract [en]

    Prognostic performance is associated with accurately estimating remaining useful life. Difficulty in accurate prognostic applications can be tackled by processing raw sensor readings into more meaningful and comprehensive health condition indicators that will then provide performance information for remaining useful life estimations. To that end, typically, multiple tasks on data pre-processing and predictions have to be carried out such that tasks can be assessed using different methodological aspects. However, incompatible methods may result in poor performance and consequently lead to undesirable error rates.

    The present research evaluates data training and prediction stages. A data-driven prognostic method based on a feed-forward neural network framework is first defined to calculate the performance of a complex system. Then, the health indicators are used in a similarity based remaining useful life estimation method. This framework presents a conceptual prognostic protocol that overcomes challenges presented by multi-regime condition monitoring data.

  • 8.
    Bektas, Oguz
    et al.
    Warwick Manufacturing Group, University of Warwick, Coventry CV4 7AL, United Kingdom.
    Jones, Jeffrey A.
    Warwick Manufacturing Group, University of Warwick, Coventry CV4 7AL, United Kingdom.
    Sankararaman, Shankar
    PricewaterhouseCoopers, San Jose, CA 95110, United States.
    Roychoudhury, Indranil
    Stinger Ghaffarian Technologies, Inc., NASA Ames Research Center, Moffett Field, CA 94035, United States.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. NASA Ames Research Center, Moffett Field, CA 94035, United States.
    Reconstructing secondary test database from PHM08 challenge data set2018In: Data in Brief, E-ISSN 2352-3409, Vol. 21, p. 2464-2469Article in journal (Refereed)
    Abstract [en]

    In this data article, a reconstructed database, which provides information from PHM08 challenge data set, is presented. The original turbofan engine data were from the Prognostic Center of Excellence (PCoE) of NASA Ames Research Center (Saxena and Goebel, 2008), and were simulated by the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) (Saxena et al., 2008). The data set is further divided into "training", "test" and "final test" subsets. It is expected from collaborators to train their models using “training” data subset, evaluate the Remaining Useful Life (RUL) prediction performance on “test” subset and finally, apply the models to the “final test” subset for competition. However, the "final test" results can only be submitted once by email to PCoE. Before the results are sent for performance evaluation, in order to pre-validate the dataset with true RUL values, this data article introduces reconstructed secondary datasets derived from the noisy degradation patterns of original trajectories. Reconstructed database refers to data that were collected from the training trajectories. Fundamentally, it is formed of individual partial trajectories in which the RUL is known as a ground truth. Its use provides a robust validation of the model developed for the PHM08 data challenge that would otherwise be ambiguous due to the high-risk of one-time submission. These data and analyses support the research data article “A Neural Network Filtering Approach for Similarity-Based Remaining Useful Life Estimations” (Bektas et al., 2018).

    Download full text (pdf)
    fulltext
  • 9.
    Chao, Manuel Arias
    et al.
    Chair of Intelligent Maintenance Systems, ETH Zurich, Switzerland.
    Kulkarni, Chetan
    SGT, Inc., NASA Ames Research Center, USA.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics.
    Fink, Olga
    Chair of Intelligent Maintenance Systems, ETH Zurich, Switzerland.
    Hybrid deep fault detection and isolation: Combining deep neural networks and system performance models2019In: International Journal of Prognostics and Health Management, E-ISSN 2153-2648, Vol. 10, no 11, article id UNSP 033Article in journal (Refereed)
    Abstract [en]

    With the increased availability of condition monitoring data on the one hand and the increased complexity of explicit system physics-based models on the other hand, the application of data-driven approaches for fault detection and isolation has recently grown. While detection accuracy of such approaches is generally very good, their performance on fault isolation often suffers from the fact that fault conditions affect a large portion of the measured signals thereby masking the fault source. To overcome this limitation, we propose a hybrid approach combining physical performance models with deep learning algorithms. Unobserved process variables are inferred with a physics-based performance model to enhance the input space of a data-driven diagnostics model. The resulting increased input space gains representation power enabling more accurate fault detection and isolation. To validate the effectiveness of the proposed method, we generate a condition monitoring dataset of an advanced gas turbine during flight conditions under healthy and four faulty operative conditions based on the Aero-Propulsion System Simulation (C-MAPSS) dynamical model. We evaluate the performance of the proposed hybrid methodology in combination with two different deep learning algorithms: deep feed forward neural networks and Variational Autoencoders, both of which demonstrate a significant improvement when applied within the hybrid fault detection and diagnostics framework. The proposed method is able to outperform pure data-driven solutions, particularly for systems with a high variability of Manuel Arias Chao et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 United States License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. operating conditions. It provides superior results both for fault detection as well as for fault isolation. For the fault isolation task, it overcomes the smearing effect that is commonly observed in pure data-driven approaches and enables a precise isolation of the affected signal. We also demonstrate that deep learning algorithms provide a better performance on the fault detection tas

  • 10.
    Chao, Manuel Arias
    et al.
    Intelligent Maintenance Systems, ETH Zurich, Switzerland.
    Kulkarni, Chetan
    SGT, Inc., NASA Ames Research Center, USA.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics.
    Fink, Olga
    Intelligent Maintenance Systems, ETH Zurich, Switzerland.
    Hybrid deep fault detection and isolation: Combining deep neural networks and system performance models2019In: International Journal of Prognostics and Health Management, E-ISSN 2153-2648, Vol. 10, no 4, article id 033Article in journal (Refereed)
    Abstract [en]

    With the increased availability of condition monitoring data on the one hand and the increased complexity of explicit system physics-based models on the other hand, the application of data-driven approaches for fault detection and isolation has recently grown. While detection accuracy of such approaches is generally very good, their performance on fault isolation often suffers from the fact that fault conditions affect a large portion of the measured signals thereby masking the fault source. To overcome this limitation, we propose a hybrid approach combining physical performance models with deep learning algorithms. Unobserved process variables are inferred with a physics-based performance model to enhance the input space of a data-driven diagnostics model. The resulting increased input space gains representation power enabling more accurate fault detection and isolation. To validate the effectiveness of the proposed method, we generate a condition monitoring dataset of an advanced gas turbine during flight conditions under healthy and four faulty operative conditions based on the Aero-Propulsion System Simulation (C-MAPSS) dynamical model. We evaluate the performance of the proposed hybrid methodology in combination with two different deep learning algorithms: deep feed forward neural networks and Variational Autoencoders, both of which demonstrate a significant improvement when applied within the hybrid fault detection and diagnostics framework. The proposed method is able to outperform pure data-driven solutions, particularly for systems with a high variability of operating conditions. It provides superior results both for fault detection as well as for fault isolation. For the fault isolation task, it overcomes the smearing effect that is commonly observed in pure data-driven approaches and enables a precise isolation of the affected signal. We also demonstrate that deep learning algorithms provide a better performance on the fault detection task compared to the traditional machine learning algorithms.

  • 11.
    Goebel, Kai
    et al.
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center, Palo Alto, CA 94304, USA.
    Rajamani, Ravi
    drR2 consulting, West Hartford, CT 06117, USA; University of Connecticut, Storrs, CT 06269, USA.
    Policy, Regulations and Standards in Prognostics and Health Management2021In: International Journal of Prognostics and Health Management, E-ISSN 2153-2648, Vol. 12, no 1Article in journal (Refereed)
    Abstract [en]

    As the field of PHM matures, it needs to be aware of the regulations, policies, and standards that will both impose boundaries as well as provide guidance for operations. All three - regulations, policies, and standards - provide information on how to design or operate something, but with different degrees of enforceability. Policies include both public policies as well as organizational policies. Operators may be required to adhere to public policies (say, an environmental policy which provides guidance for the pollution prevention act (the latter is a US law)) whereas organisational policies often reflect strategic considerations within private organizations (such as maintenance policies). Regulations (such as aeronautics or nuclear energy) typically impose binding rules of engagement and are imposed by regulatory bodies that are responsible for a particular field. Standards, in contrast, are community-consensus guidelines that are meant to provide benefit to the community by describing best practices. Adoption of such guidelines is entirely voluntary but may provide benefits by not having to reinvent the wheel and for finding common ground amongst other adopters. Awareness of both guidelines and barriers will enable practitioners in adopting best practices within the legal constraints. This paper provides an overview of the current regulations, policies, and standards in the field of Prognostics and Health Management.

  • 12.
    Goebel, Kai
    et al.
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center, Palo Alto, CA, United States.
    Smith, Brian
    NASA Ames Research Center, Moffett Field, CA, United States.
    Bajawa, Anupa
    NASA Ames Research Center, Moffett Field, CA, United States.
    Ethics in prognostics and health management2019In: International Journal of Prognostics and Health Management, E-ISSN 2153-2648, Vol. 10, no 1, article id 012Article in journal (Refereed)
    Abstract [en]

    As we are entering an era where intelligent systems are omnipresent and where they also penetrate Prognostics and Health Management (PHM), the discussion of moral machines or ethics in engineering will inevitably engulf PHM as well. This article explores the topic of ethics within the PHM domain: how it is relevant, and how it may be dealt with in a conscientious way. The paper provides a historical perspective on ethics-related developments that resulted in the formulation of engineering ethics codes, regulations, and policies. By virtue of these developments, ethics has already been encapsulated in PHM systems. The specific areas that have traditionally driven ethics considerations include safety and security, and they increasingly include privacy, and environmental protection. During the course of future technology development, innovations will increasingly impact all of these topics. It is argued that consciously embracing these issues will increase the competitive advantage of a PHM technology solution. As a guideline, specific ethics attributes are derived from professional engineering ethics codes, and a path towards insertion into a requirements flowdown is suggested.

  • 13.
    Hu, Jueming
    et al.
    Arizona State University, School for Engineering of Matter, Transport & Energy, 501 E Tyler Mall, ENGRC 419, Tempe, AZ 85281, United States.
    Erzberger, Heinz
    University of California, Santa Cruze, 95064, United States.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics.
    Liu, Yongming
    Arizona State University, School for Engineering of Matter, Transport & Energy, 501 E Tyler Mall, ENGRC 419, Tempe, AZ 85281, United States.
    Conflict Probability Estimation Using a Risk-Based Dynamic Anisotropic Operational Safety Bound for UAS Traffic Management2020In: AIAA Scitech 2020 Forum, American Institute of Aeronautics and Astronautics, 2020, article id 2020-0738Conference paper (Refereed)
    Abstract [en]

    The safety and efficiency of rotary-wing UAV traffic management require an operational safety bound and fast conflict prediction. A novel method to determine probabilistic risk-based operational safety bound for rotary-wing UAV traffic management is proposed. The key idea is to include probabilistic uncertainty quantification of the safety bound. The unique design for the operational safety bound results in a dynamic and anisotropic shape of the bound which considers the vectorized velocity of the UAV and wind. Operational safety bound is used to identify a virtual geographic boundary to protect aircraft and to ensure airspace safety. The proposed operational safety bound is calculated as a function of vehicle performance characteristics, state of vehicle, wind, and other probabilistic parameters that affect the real position of vehicle, such as the position error from the Global Positioning System (GPS). This paper presents an efficient method to estimate the probability that a conflict will occur between a UAV pair in confined airspace using the proposed risk-based dynamic anisotropic operational safety bounds. Conflict probability is critical to evaluate airspace capacity and to determine the optimal time to initiate conflict resolution maneuver. Several conclusions and suggestions of further research directions are given.

  • 14.
    Hu, Jueming
    et al.
    Arizona State University, Tempe, Arizona 85281.
    Erzberger, Heinz
    University of California, Santa Cruz, California 95064.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center, Palo Alto, California 94304.
    Liu, Yongming
    Arizona State University, Tempe, Arizona 85281.
    Probabilistic Risk-Based Operational Safety Bound for Rotary-Wing Unmanned Aircraft Systems Traffic Management2020In: Journal of Aerospace Information Systems, ISSN 1940-3151, Vol. 17, no 3, p. 171-181Article in journal (Refereed)
    Abstract [en]

    A novel method to determine probabilistic operational safety bound for rotary-wing unmanned aircraft systems (UAS) traffic management is proposed in this paper. The key idea is to combine a deterministic model for rotary-wing UAS flying distance estimation to avoid conflict and a probabilistic uncertainty quantification methodology to evaluate the risk level (defined as the probability of failure) of separation loss between UAS. The proposed methodology results in a dynamic and probabilistic airspace reservation to ensure the safety and efficiency for future UAS operations. The model includes UAS performance, system updating frequency and accuracy, and weather conditions. Also, the parameterized probabilistic model includes various uncertainties from different sources and develops an anisotropic operational safety bound. Monte Carlo simulations are used to illustrate the operational safety bound determination with a specified risk level (i.e., probability of failure). It is known that uncertainty plays an important role in determining the operational safety bound size, and the proposed methodology provides a simple and efficient quantification of uncertainty impact on the safety bound with a prescribed risk level. It is also providing a useful tool to quantify uncertainty reduction with additional information and measurements in future UAS operations.

  • 15.
    Hu, Jueming
    et al.
    Arizona State University, Tempe, AZ, 85281.
    Erzberger, Heinz
    University of California, Santa Cruz, 95064.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. PARC, Palo Alto, CA, 94304.
    Liu, Yongming
    Arizona State University, Tempe, AZ, 85281.
    Risk-Based Dynamic Anisotropic Operational Safety Bound for Rotary UAV Traffic Control2019In: Proceedings of the Annual Conference of the Prognostics and Health Management Society 2019, Prognostics and Health Management Society, PHM , 2019Conference paper (Other academic)
    Abstract [en]

    This paper proposed a novel method to determine probabilistic operational safety bound for unmanned aircraft traffic management. The key idea is to implement probabilistic uncertainty quantification and design the operational safety bound shape considering UAV’s heading direction. Operational safety bound is used to identify a virtual geographic boundary to protect aircraft and to ensure airspace safety. The proposed operational safety bound is calculated as a function of vehicle performance characteristics, state of vehicle, weather and other probabilistic parameters that affect the real position of vehicle such as position error from the Global Positioning System (GPS). It is calculated individually for each vehicle using real-time data and probability simulation. It considers the heading direction of vehicle and thus it is an anisotropic design. Monte Carlo simulations are conducted to estimate the operational safety bound size with a specified probability of failure. Results indicate that uncertainty is crucial for the operational safety bound’s size. Sensitivity study shows that UAV speed has the largest effect on the operational safety bound size. Analysis of impact of failure probability shows that operational safety bound size increases with the decrease in allowable failure probability, but the bound size based on different operational safety bound concept increases at different rate.

  • 16.
    Hulse, Daniel
    et al.
    Oregon State University, Corvallis, Oregon 97330.
    Biswas, Arpan
    Oregon State University, Corvallis, Oregon 97330.
    Hoyle, Christopher
    Oregon State University, Corvallis, Oregon 97330.
    Tumer, Irem Y.
    Oregon State University, Corvallis, Oregon 97330.
    Kulkarni, Chetan
    KBR, Inc., NASA Ames Research Center, Moffett Field, California 94035.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center, Palo Alto, California 94304.
    Exploring Architectures for Integrated Resilience Optimization2021In: Journal of Aerospace Information Systems, E-ISSN 2327-3097, Vol. 18, no 10, p. 665-678Article in journal (Refereed)
  • 17.
    Hulse, Daniel
    et al.
    Oregon State University, Corvallis, Oregon, 97330, USA.
    Hoyle, Christopher
    Oregon State University, Corvallis, Oregon, 97330, USA.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center, Palo Alto, California 94304, USA.
    Tumer, Irem
    Oregon State University, Corvallis, Oregon, 97330, USA.
    Using Value Assessment to Drive PHM System Development in Early Design2019In: Proceedings of the Annual Conference of the Prognostics and Health Management Society 2019, Prognostics and Health Management Society , 2019Conference paper (Other academic)
    Abstract [en]

    Prognostics and Health Management (PHM) systems have been shown to provide many benefits to the reliability, performance, and life of engineered systems. However, because of trade-offs between up-front design and implementation costs, operational performance, and reliability, it may not be obvious in the early design phase whether one PHM system will be more beneficial to another, or whether a PHM system will provide benefit compared to a traditional reliability approach. These trade-offs make the commitment required to pursue PHM features in the early design phase difficult to justify. In this paper, a cost model incorporating trade-offs among design cost, operational performance, and failure risk is used to provide a comprehensive value comparison of health management options to motivate design decision-making. This approach is then demonstrated in a simple case study comparing the use of a PHM system for condition-based maintenance or diagnostic-based recovery with implementing redundancy and increased inspection in the design. Then it is shown how different model inputs and assumptions result in a different system value (and different design choice from the process), illustrating the usefulness of cost modelling to capture design trade-offs. Using this approach, decisions about pursuing PHM can be made early, enabling the benefits to be fully leveraged in the design process to achieve increased operational resilience.

  • 18.
    Hulse, Daniel
    et al.
    School of Mechanical, Industrial and Manufacturing Engineering, Oregon State University, Corvallis, Oregon, USA.
    Hoyle, Christopher
    School of Mechanical, Industrial and Manufacturing Engineering, Oregon State University, Corvallis, Oregon, USA.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Discovery and Systems Health, Intelligent Systems Division, NASA Ames Research Center, Moffett Field, California, USA.
    Tumer, Irem Y.
    School of Mechanical, Industrial and Manufacturing Engineering, Oregon State University, Corvallis, Oregon, USA.
    Quantifying the Resilience-Informed Scenario Cost Sum: A Value-Driven Design Approach for Functional Hazard Assessment2019In: Journal of mechanical design (1990), ISSN 1050-0472, E-ISSN 1528-9001, Vol. 141, no 2, article id MD-18-1503Article in journal (Refereed)
    Abstract [en]

    Complex engineered systems can carry risk of high failure consequences, and as a result, resilience—the ability to avoid or quickly recover from faults—is desirable. Ideally, resilience should be designed-in as early in the design process as possible so that designers can best leverage the ability to explore the design space. Toward this end, previous work has developed functional modeling languages which represent the functions which must be performed by a system and function-based fault modeling frameworks have been developed to predict the resulting fault propagation behavior of a given functional model. However, little has been done to formally optimize or compare designs based on these predictions, partially because the effects of these models have not been quantified into an objective function to optimize. The work described herein closes this gap by introducing the resilience-informed scenario cost sum (RISCS), a scoring function which integrates with a fault scenario-based simulation, to enable the optimization and evaluation of functional model resilience. The scoring function accomplishes this by quantifying the expected cost of a design's fault response using probability information, and combining this cost with design and operational costs such that it may be parameterized in terms of designer-specified resilient features. The usefulness and limitations of using this approach in a general optimization and concept selection framework are discussed in general, and demonstrated on a monopropellant system design problem. Using RISCS as an objective for optimization, the algorithm selects the set of resilient features which provides the optimal trade-off between design cost and risk. For concept selection, RISCS is used to judge whether resilient concept variants justify their design costs and make direct comparisons between different model structures.

  • 19.
    Hulse, Daniel
    et al.
    School of Mechanical, Industrial and Manufacturing Engineering, Oregon State University, Corvallis, OR 97330.
    Hoyle, Christopher
    School of Mechanical, Industrial and Manufacturing Engineering, Oregon State University, Corvallis, OR 97330.
    Tumer, Irem Y.
    School of Mechanical, Industrial and Manufacturing Engineering, Oregon State University, Corvallis, OR 97330.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center, Palo Alto, CA 94304.
    How Uncertain Is Too Uncertain? Validity Tests for Early Resilient and Risk-Based Design Processes2021In: Journal of mechanical design (1990), ISSN 1050-0472, E-ISSN 1528-9001, Vol. 143, no 1, article id 011702Article in journal (Refereed)
    Abstract [en]

    A number of risk and resilience-based design methods have been put forward over the years that seek to provide designers the tools to reduce the effects of potential hazards in the early design phase. However, because of the associated high level of uncertainty and low-fidelity design representations, one might justifiably wonder if using a resilient design process in the early design phase will reliably produce useful results that would improve the realized design. This paper provides a testing framework for design processes that determines the validity of the process by quantifying the epistemic uncertainty in the assumptions used to make decisions. This framework uses this quantified uncertainty to test whether three metrics are within desirable bounds: the change in the design when uncertainty is considered, the increase in the expected value of the design, and the cost of choice-related uncertainty. This approach is illustrated using two examples to demonstrate how both discrete and continuous parametric uncertainty can be considered in the testing procedure. These examples show that early design process validity is sensitive to the level of uncertainty and magnitude of design changes, suggesting that while there is a justifiable decision-theoretic case to consider high-level, high-impact design changes during the early design phase, there is less of a case to choose between relatively similar design options because the cost of making the choice under high uncertainty is greater than the expected value improvement from choosing the better design.

  • 20.
    Hulse, Daniel
    et al.
    Department of M.I.M.E., Oregon State University, Corvallis, OR, 97330, United States.
    Walsh, Hannah
    Department of M.I.M.E., Oregon State University, Corvallis, OR, 97330, United States.
    Dong, Andy
    Department of M.I.M.E., Oregon State University, Corvallis, OR, 97330, United States.
    Hoyle, Christopher
    Department of M.I.M.E., Oregon State University, Corvallis, OR, 97330, United States.
    Tumer, Irem
    Department of M.I.M.E., Oregon State University, Corvallis, OR, 97330, United States.
    Kulkarni, Chetan
    KBR, Inc., NASA Ames Research Center, Moffett Field, CA, 94035, United States.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics.
    Fmdtools: A Fault Propagation Toolkit for Resilience Assessment in Early Design2021In: International Journal of Prognostics and Health Management, E-ISSN 2153-2648, Vol. 12, no 3, p. 1-18Article in journal (Refereed)
    Abstract [en]

    Incorporating resilience in design is important for the long-term viability of complex engineered systems. Complex aerospace systems, for example, must ensure safety in the event of hazards resulting from part failures and external circumstances while maintaining efficient operations. Traditionally, mitigating hazards in early design has involved experts manually creating hazard analyses in a time-consuming process that hinders one’s ability to compare designs. Furthermore, as opposed to reliability-based design, resilience-based design requires using models to determine the dynamic effects of faults to compare recovery schemes. Models also provide design opportunities, since models can be parameterized and optimized and because the resulting hazard analyses can be updated iteratively. While many theoretical frameworks have been presented for early hazard assessment, most currently-available modelling tools are meant for the later stages of design. Given the wide adoption of Python in the broader research community, there is an opportunity to create an environment for researchers to study the resilience of different PHM technologies in the early phases of design. This paper describes fmdtools, an attempt to realize this opportunity with a set of modules which may be used to construct different design models, simulate system behaviors over a set of fault scenarios and analyze the resilience of the resulting simulation results. This approach is demonstrated in the hazard analysis and architecture design of a multi-rotor drone, showing how the toolkit enables a large number of analyses to be performed on a relatively simple model as it progresses through the early design process.

  • 21.
    Li, Z.
    et al.
    Department of Mechanical and Aerospace Engineering, University of Central Florida, Orlando, United States.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. NASA Ames Research Center, Moffett Field, United States.
    Wu, D.
    Department of Mechanical and Aerospace Engineering, University of Central Florida, Orlando, United States.
    Degradation Modeling and Remaining Useful Life Prediction of Aircraft Engines Using Ensemble Learning2019In: Journal of engineering for gas turbines and power, ISSN 0742-4795, E-ISSN 1528-8919, Vol. 41, no 4, article id 041008Article in journal (Refereed)
    Abstract [en]

    Degradation modeling and prediction of remaining useful life (RUL) are crucial to prognostics and health management of aircraft engines. While model-based methods have been introduced to predict the RUL of aircraft engines, little research has been reported on estimating the RUL of aircraft engines using novel data-driven predictive modeling methods. The objective of this study is to introduce an ensemble learning-based prognostic approach to modeling an exponential degradation process due to wear as well as predicting the RUL of aircraft engines. The ensemble learning algorithm combines multiple base learners, including random forests (RFs), classification and regression tree (CART), recurrent neural networks (RNN), autoregressive (AR) model, adaptive network-based fuzzy inference system (ANFIS), relevance vector machine (RVM), and elastic net (EN), to achieve better predictive performance. The particle swarm optimization (PSO) and sequential quadratic optimization (SQP) methods are used to determine optimum weights that are assigned to the base learners. The predictive model trained by the ensemble learning algorithm is demonstrated on the data generated by the commercial modular aero-propulsion system simulation (C-MAPSS) tool. Experimental results have shown that the ensemble learning algorithm predicts the RUL of the aircraft engines with considerable robustness as well as outperforms other prognostic methods reported in the literature. 

  • 22.
    Matelli, José Alexandre
    et al.
    NASA Ames Research Center, Intelligent Systems Division, Discovery and Systems Health, Moffett Field, CA.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. NASA Ames Research Center, Intelligent Systems Division, Discovery and Systems Health.
    Conceptual design of cogeneration plants under a resilient design perspective: Resilience metrics and case study2018In: Applied Energy, ISSN 0306-2619, E-ISSN 1872-9118, Vol. 215, p. 736-750Article in journal (Refereed)
    Abstract [en]

    The conceptual design phase is the first step in the design process of an engineering system. Most engineering systems, including cogeneration plants, may and likely will experience some malfunctions during its life cycle. The metrics typically considered in the conceptual design phase (and for analysis and optimization) of energy systems are cost, efficiency and environmental impacts. Quite rarely are operational considerations about malfunctions integrated during the conceptual design phase. Resilient design, or design for resilience, addresses this gap as illustrated here in the area of energy conversion and conservation of energy processes by examining the conceptual design of a cogeneration plant. Resilient design is a relatively new research field where the engineering system is designed such that it can optimally recover from failures. The main challenge is to quantify the resilience in early design phases, since there is not much detailed information about system components available at this point. To address these challenges, this paper introduces a novel resilient design framework that uses new metrics within a Monte Carlo-based assessment approach. The framework is exercised on conceptual designs of cogeneration plants. Results from this framework are compared against those from a methodology based on complex networks theory that has been previously suggested in the literature. The former presented more consistent results than the latter and we discuss the differences. Results also show that the concept with higher efficiency was not the one with higher resilience. Finally, we discuss how to integrate specific failure probabilities information into the framework (should that information be available), and deliberate on relations between resilience, fault handling strategies and design requirements.

  • 23.
    Matelli, José Alexandre
    et al.
    São Paulo State University (UNESP), School of Engineering, Department of Energy, Av. Ariberto Pereira da Cunha, Guaratinguetá, SP, Brazil.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. NASA Ames Research Center, Intelligent Systems Division, Discovery and Systems Health, Moffett Field, CA.
    Resilience evaluation of the environmental control and life support system of a spacecraft for deep space travel2018In: Acta Astronautica, ISSN 0094-5765, E-ISSN 1879-2030, Vol. 152, p. 360-369Article in journal (Refereed)
    Abstract [en]

    In deep space manned travels, the crew life will be totally dependent on the environment control and life support system of the spacecraft. A life-support system for manned missions is a set of technologies to regenerate the basic life-support elements, such as oxygen and water, which makes resilience a paramount feature of this system. The resilience of a complex engineered system is the ability of the system to withstand failures, continue operating and recover from those failures with minimum disruption. Resilient design is a new design framework on which the main goal is to quantify system resilience upfront in order to guide the design team during the conceptual design stage. In this article, we present a tool that combines a rule-based approach with a Monte Carlo-based approach to evaluate the resilience of a proposed environment control and life support system designed for deep space travel. Based on the results found, we explore a few design alternatives in order to increase system resilience.

  • 24.
    Mishra, Madhav
    et al.
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. SKF-LTU University Technology Centre.
    Martinsson, Jesper
    Luleå University of Technology, Department of Engineering Sciences and Mathematics, Mathematical Science.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. NASA Ames Research Center, Intelligent Systems Division, Moffett Field, CA. USA.
    Rantatalo, Matti
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics.
    Bearing Life Prediction with Informed Hyperprior Distribution: A Bayesian Hierarchical and Machine Learning Approach2021In: IEEE Access, E-ISSN 2169-3536, Vol. 9, p. 157002-157011Article in journal (Refereed)
    Abstract [en]

    A Bayesian hierarchical model (BHM) is developed to predict bearing life using envelope acceleration data in combination with a degradation model and prior knowledge of the bearing rating life. The BHM enables the inference of individual bearings, groups of bearings, or bearings operating under certain conditions. The key benefit of the BHM approach is that the relationships between the bearing model parameters and their prior distributions can be expressed at different hierarchical levels. We begin our analysis using a bearing rating life calculation L10h and an estimate of its associated failure time distribution. Realistic variations to constrain our prior distribution of the failure time are then applied before measurements are available. When data become available, estimates more representative of our specific batch and operating conditions are inferred, both on the individual bearing level and the bearing group level. The proposed prognostics methodology can be used in situations with varying amounts of data. The presented BHM approach can also be used to predict the remaining useful life (RUL) of bearings both in situations in which the bearing is considered to be in a healthy state and in situations after a defect has been detected.

  • 25.
    Shi, Junchuan
    et al.
    Department of Mechanical and Aerospace Engineering, University of Central Florida, Orlando, FL 32816, USA.
    Peng, Dikang
    School of Mechanical and Manufacturing Engineering, University of New South Wales, Sydney, NSW 2052, Australia.
    Peng, Zhongxiao
    School of Mechanical and Manufacturing Engineering, University of New South Wales, Sydney, NSW 2052, Australia.
    Zhang, Ziyang
    Department of Mechanical and Aerospace Engineering, University of Central Florida, Orlando, FL 32816, USA.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center, Palo Alto, CA 94034, USA.
    Wu, Dazhong
    Department of Mechanical and Aerospace Engineering, University of Central Florida, Orlando, FL 32816, USA.
    Planetary gearbox fault diagnosis using bidirectional-convolutional LSTM networks2022In: Mechanical systems and signal processing, ISSN 0888-3270, E-ISSN 1096-1216, Vol. 162, article id 107996Article in journal (Refereed)
    Abstract [en]

    Gearbox fault diagnosis is expected to significantly improve the reliability, safety and efficiency of power transmission systems. However, planetary gearbox fault diagnosis remains a challenge due to complex responses caused by multiple planetary gears. Model-based gearbox fault diagnosis techniques extract hand-crafted features from sensor data based on underlying physics and statistical analysis, which are not effective in extracting spatial and temporal features automatically. While deep learning methods such as convolutional neural network (CNN) enable automatic feature extraction from multiple sensor sources, they are not capable of extracting spatial and temporal features simultaneously without losing critical feature information. To address this issue, we introduce a novel deep neural network based on bidirectional-convolutional long short-term memory (BiConvLSTM) networks to determine the type, location, and direction of planetary gearbox faults by extracting spatial and temporal features from both vibration and rotational speed measurements automatically and simultaneously. In particular, a CNN determines spatial correlations between two measurements within one time step automatically by combining signals collected from three accelerometers and one tachometer. Long short-term memory (LSTM) networks identify temporal dependencies between two adjacent time steps. By replacing input-to-state and state-to-state operations in the LSTM cell with convolutional operations, the BiConvLSTM can learn spatial correlations and temporal dependencies without losing critical features. Experimental results have shown that the BiConvLSTM network can detect the type, location, and direction of gearbox faults with higher accuracy than conventional deep learning approaches such as CNN, LSTM, and CNN-LSTM.

  • 26.
    Shi, Junchuan
    et al.
    Department of Mechanical and Aerospace Engineering, University of Central Florida, Orlando, FL 32816.
    Yu, Tianyu
    Department of Mechanical and Aerospace Engineering, University of Central Florida, Orlando, FL 32816.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center, Palo Alto, CA 94034. NASA Ames Research Center, Moffett Field, CA 94035.
    Wu, Dazhong
    Department of Mechanical and Aerospace Engineering, University of Central Florida, Orlando, FL 32816.
    Remaining Useful Life Prediction of Bearings Using Ensemble Learning: The Impact of Diversity in Base Learners and Features2021In: Journal of Computing and Information Science in Engineering, ISSN 1530-9827, E-ISSN 1944-7078, Vol. 21, no 2, article id 021004Article in journal (Refereed)
    Abstract [en]

    Prognostics and health management (PHM) of bearings is crucial for reducing the risk of failure and the cost of maintenance for rotating machinery. Model-based prognostic methods develop closed-form mathematical models based on underlying physics. However, the physics of complex bearing failures under varying operating conditions is not well understood yet. To complement model-based prognostics, data-driven methods have been increasingly used to predict the remaining useful life (RUL) of bearings. As opposed to other machine learning methods, ensemble learning methods can achieve higher prediction accuracy by combining multiple learning algorithms of different types. The rationale behind ensemble learning is that higher performance can be achieved by combining base learners that overestimate and underestimate the RUL of bearings. However, building an effective ensemble remains a challenge. To address this issue, the impact of diversity in base learners and extracted features in different degradation stages on the performance of ensemble learning is investigated. The degradation process of bearings is classified into three stages, including normal wear, smooth wear, and severe wear, based on the root-mean-square (RMS) of vibration signals. To evaluate the impact of diversity on prediction performance, vibration data collected from rolling element bearings was used to train predictive models. Experimental results have shown that the performance of the proposed ensemble learning method is significantly improved by selecting diverse features and base learners in different degradation stages.

  • 27.
    Sierra, G.
    et al.
    Department of Electrical Engineering, University of Chile, Santiago.
    Orchard, M.
    Department of Electrical Engineering, University of Chile, Santiago, Chile.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. NASA Ames Research Center, Moffett Field, CA, USA.
    Kulkarni, C.
    SGT Inc., NASA Ames Research Center, Moffett Field, CA, USA.
    Battery Health Management for Small-size Battery-powered Rotary-wing Unmanned Aerial Vehicles: An Efficient Approach for Constrained Computing Platforms2019In: Reliability Engineering & System Safety, ISSN 0951-8320, E-ISSN 1879-0836, Vol. 182, p. 166-178Article in journal (Refereed)
    Abstract [en]

    This article presents a holistic framework for the design, implementation and experimental validation of Battery Management Systems (BMS) in rotatory-wing Unmanned Aerial Vehicles (UAVs) that allows to accurately (i) estimate the State of Charge (SOC), and (ii) predict the End of Discharge (EOD) time of lithium-polymer batteries in small-size multirotors by using a model-based prognosis architecture that is efficient and feasible to implement in low-cost hardware. The proposed framework includes a simplified battery model that incorporates the electric load dependence, temperature dependence and SOC dependence by using the concept of Artificial Evolution to estimate some of its parameters, along with a novel Outer Feedback Correction Loop (OFCL) during the estimation stage which adjusts the variance of the process noise to diminish bias in Bayesian state estimation and helps to compensate problems associated with incorrect initial conditions in a non-observable dynamic system. Also, it provides an aerodynamic-based characterization of future power consumption profiles. A quadrotor has been used as validation platform. The results of this work will allow making decisions about the flight plan and having enough confidence in those decisions so that the mission objectives can be optimally achieved.

  • 28.
    Sierra, Gina
    et al.
    NASA Ames Research Center, Moffett Field, CA 94035, USA.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center, Palo Alto, CA 94304, USA.
    A Novel Model-Based Framework for Power Consumption Prognosis in Multirotors Applied to the Battery End of Discharge Prognostics2020In: Journal of Nondestructive Evaluation, Diagnostics and Prognostics of Engineering Systems, ISSN 2572-3901, Vol. 3, no 2Article in journal (Refereed)
    Abstract [en]

    Accurately estimating the time of battery end of discharge (EOD) in electric unmanned aerial vehicles (UAVs) provides assurance that a given mission can be completed before the energy stored in the battery runs out and aids decision-making processes such as mission replanning to mitigate shortcomings associated with the available energy. The accuracy of the predicted battery EOD time is strongly correlated to the accuracy of the expected power consumption during the mission. This paper reports on a novel model-based framework for power consumption prognosis in multirotors which includes an improved power consumption model that characterizes the power required by a multirotor in axial and nonaxial translation and incorporates the wind effects on the required power. A particle filter is used in conjunction with the concept of artificial evolution to estimate and monitor wind speed, wind direction, and thrust based on measurements of power. Monte Carlo sampling-based predictor is used to predict the trajectories of power used in battery EOD prognostic. The framework is applied to battery EOD prognostic of a quadcopter that performs a delivery mission with low horizontal speed (where rotor tilt is not a significant factor). Results show that predicted trajectories of power accurately represent the uncertainty of future power consumption. Even when certain information (such as aircraft weight) is not available at every time-step, the framework allows tracking the actual EOD time because of its capability to monitor thrust. These results demonstrate the effectiveness of the proposed framework. © 2020 American Society of Mechanical Engineers (ASME). All rights reserved.

  • 29.
    Sierra, Gina
    et al.
    SGT/KBR, Inc., NASA Ames Research Center, Moffett Field, CA 94035, USA.
    Robinson, Elinirina I.
    SGT/KBR, Inc., NASA Ames Research Center, Moffett Field, CA 94035, USA.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Palo Alto Research Center, Palo Alto, CA 94304, USA.
    Improving tail accuracy of the predicted cumulative distribution function of time of failure2021In: Reliability Engineering & System Safety, ISSN 0951-8320, E-ISSN 1879-0836, Vol. 207, article id 107333Article in journal (Refereed)
    Abstract [en]

    Prognostic information is used to make decisions such as when to perform maintenance or - in time sensitive and safety critical applications - when to change operational settings. Where distributions about expected end of life (EOL) are available, these decisions are often based on risk-informed thresholds, for example a 2σ or 3σ criterion which considers the probability of making a bad decision at 5% or 0.3%, respectively, as tolerable. Sampling-based techniques such as Monte Carlo Sampling (MCS) and Latin Hypercube Sampling (LHS) can provide effective approaches to the propagation and analysis of uncertainty. Due to its efficient manner of stratifying across the range of each sampled variable, LHS requires less computational effort than MCS and is therefore more often used. However, since the focus is placed on accurately predicting the tails of the Cumulative Distribution Function (CDF) of Time of Failure (ToF) sampling-base techniques may not properly represent these areas. Although one might be tempted to use a brute force approach and simply increase the number of samples, some safety-critical applications may be computationally constrained. Such applications include electric UAV where the decision making process has to be fast in order to take action as soon as possible. This paper explores the ability of MCS and LHS to perform tail prediction with small sample sizes. The results show that LHS does not provide a significant advantage over MCS in terms of characterizing the tails of the CDF of the battery End of Discharge (EOD) prediction. Then, a methodology that combines MCS and Kernel Density Estimation (KDE) is investigated. The advantages of KDE in terms of reducing sample size while improving tail accuracy are demonstrated on battery end-of-discharge data.

  • 30.
    Tian, Yuan
    et al.
    ETH Zürich, Switzerland.
    Arias Chao, Manuel
    ETH Zürich, Switzerland.
    Kulkarni, Chetan
    KBR Inc., United States of America; NASA Ames Research Center, United States of America.
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics.
    Fink, Olga
    ETH Zürich, Switzerland.
    Real-time model calibration with deep reinforcement learning2022In: Mechanical systems and signal processing, ISSN 0888-3270, E-ISSN 1096-1216, Vol. 165, article id 108284Article in journal (Refereed)
    Abstract [en]

    The real-time, and accurate inference of model parameters is of great importance in many scientific and engineering disciplines that use computational models (such as a digital twin) for the analysis and prediction of complex physical processes. However, fast and accurate inference for processes of complex systems cannot easily be achieved in real-time with state-of-the-art methods under noisy real-world conditions with the requirement of a real-time response. The primary reason is that the inference of model parameters with traditional techniques based on optimization or sampling often suffers from computational and statistical challenges, resulting in a trade-off between accuracy and deployment time. In this paper, we propose a novel framework for inference of model parameters based on reinforcement learning. The proposed methodology is demonstrated and evaluated on two different physics-based models of turbofan engines. The experimental results demonstrate that the proposed methodology outperforms all other tested methods in terms of speed and robustness, with high inference accuracy.

  • 31.
    Zhang, Xiaoge
    et al.
    Department of Civil and Environmental Engineering, Vanderbilt University, Nashville, USA.
    Mahadevan, Sankaran
    Department of Civil and Environmental Engineering, Vanderbilt University, Nashville, USA..
    Goebel, Kai
    Luleå University of Technology, Department of Civil, Environmental and Natural Resources Engineering, Operation, Maintenance and Acoustics. Systems Sciences Lab, Palo Alto Research Center, Palo Alto, CA, USA.
    Network Reconfiguration for Increasing Transportation System Resilience Under Extreme Events2019In: Risk Analysis, ISSN 0272-4332, E-ISSN 1539-6924, Vol. 39, no 9, p. 2054-2075Article in journal (Refereed)
    Abstract [en]

    Evacuating residents out of affected areas is an important strategy for mitigating the impact of natural disasters. However, the resulting abrupt increase in the travel demand during evacuation causes severe congestions across the transportation system, which thereby interrupts other commuters' regular activities. In this article, a bilevel mathematical optimization model is formulated to address this issue, and our research objective is to maximize the transportation system resilience and restore its performance through two network reconfiguration schemes: contraflow (also referred to as lane reversal) and crossing elimination at intersections. Mathematical models are developed to represent the two reconfiguration schemes and characterize the interactions between traffic operators and passengers. Specifically, traffic operators act as leaders to determine the optimal system reconfiguration to minimize the total travel time for all the users (both evacuees and regular commuters), while passengers act as followers by freely choosing the path with the minimum travel time, which eventually converges to a user equilibrium state. For each given network reconfiguration, the lower-level problem is formulated as a traffic assignment problem (TAP) where each user tries to minimize his/her own travel time. To tackle the lower-level optimization problem, a gradient projection method is leveraged to shift the flow from other nonshortest paths to the shortest path between each origin-destination pair, eventually converging to the user equilibrium traffic assignment. The upper-level problem is formulated as a constrained discrete optimization problem, and a probabilistic solution discovery algorithm is used to obtain the near-optimal solution. Two numerical examples are used to demonstrate the effectiveness of the proposed method in restoring the traffic system performance.

1 - 31 of 31
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf