George Vachtsevanos - Academia.edu (original) (raw)

Papers by George Vachtsevanos

Fault diagnosis and failure prognosis of critical dynamic systems, such as aircraft and industria... more Fault diagnosis and failure prognosis of critical dynamic systems, such as aircraft and industrial processes, rely on degradation or fatigue models and measurements typically acquired on-line in real-time. Such sensor data must be pre-processed in order to remove artifacts and improve the signal-to-noise ratio. Furthermore, they must be processed appropriately so that useful information in compact form can be extracted and used to detect incipient failures and predict the remaining useful life of failing components. We present a methodology to select an optimum feature vector from a list of candidate features, prioritize and rank them to meet set performance objectives. The enabling technologies include genetic programming tools, data fusion and model-based approaches for feature selection and extraction. We will suggest a multi-core processing environment for the efficient and expedient implementation of these technologies. Performance metrics are defined to assess the efficacy of the methodology. Typical examples from aircraft systems are used to demonstrate the proposed techniques.

Proceedings of the ... American Control Conference, Jul 1, 2007

Critical aircraft assets are required to be available when needed, while exhibiting attributes of... more Critical aircraft assets are required to be available when needed, while exhibiting attributes of reliability, robustness, and high confidence under a variety of flight regimes, and maintained on the basis of their current condition rather than on the basis of scheduled maintenance practices. New and innovative technologies must be developed and implemented to address these concerns. Condition-based maintenance requires that the health of critical components/systems be monitored and diagnostic/prognostic strategies be developed to detect and identify incipient failures and predict the failing component's remaining useful life. Typically, vibration and other key indicators onboard an aircraft are severely corrupted by noise, thus curtailing the ability to accurately diagnose and predict failures. This paper introduces a novel blind deconvolution denoising scheme that employs a vibration model in the frequency domain and attempts to arrive at the true vibration signal through an iterative optimization process. Performance indexes are defined and data from a helicopter are used to demonstrate the effectiveness of the proposed approach. Index Terms-Blind deconvolution, planetary gear train, vibration signal denoising. I. INTRODUCTION F AULT DETECTION, isolation, and identification, especially for safety critical components and subsystems, have recently drawn increasing interest in the condition-based maintenance (CBM) community [1]-[4]. Epicyclic, or planetary, gear trains are widely used in the main transmission of many systems, such as the helicopter and other aircraft [5], [7]. This kind of gear system consists of an inner sun gear, two or more rotating planet gears, a stationary outer ring gear, and a planetary carrier. The planet gears, which surround the sun gear, are riding on the planetary carrier and also rotating inside the outer ring gear. In the operation, torque is transmitted through the sun gear to the planets and planetary carrier. The carrier, in turn, transmits torque to the main rotor shaft and blades [6]-[9]. The gearbox is a critical component that directly impacts the safety and performance of the aircraft.

IEEE Transactions on Biomedical Engineering, Aug 1, 2003

Springer eBooks, 2009

Mechanisms underlying seizure generation are traditionally thought to act over seconds to minutes... more Mechanisms underlying seizure generation are traditionally thought to act over seconds to minutes before clinical seizure onset. We analyzed continuous 3- to 14-day intracranial EEG recordings from five patients with mesial temporal lobe epilepsy obtained during evaluation for epilepsy surgery. We found localized quantitative EEG changes identifying prolonged bursts of complex epileptiform discharges that became more prevalent 7 hr before seizures and highly localized subclinical seizure-like activity that became more frequent 2 hr prior to seizure onset. Accumulated energy increased in the 50 min before seizure onset, compared to baseline. These observations, from a small number of patients, suggest that epileptic seizures may begin as a cascade of electrophysiological events that evolve over hours and that quantitative measures of preseizure electrical activity could possibly be used to predict seizures far in advance of clinical onset.

Annual Conference of the PHM Society, Oct 2, 2014

With electrical power supplies playing an important role in the operation of aircraft systems and... more With electrical power supplies playing an important role in the operation of aircraft systems and subsystems , flight and ground crews need health state awareness and prediction tools that accurately diagnose faults, predict failures, and project remaining life of these onboard power supplies. Among onboard power supplies, switch-mode power supplies are commonly used where their weight, size, and efficiency make them preferable to conventional transformer-based power supplies. In this paper, we present a framework of diagnostics and prognostics methodology based on an equivalent circuit system simulation model developed from a commercially available switch-mode power supply, and empirical component degradation models. In industrial applications, case-specified modifications can be made according to specific experimental or service conditions of different commercial products. First, the developed simulation model is validated through experimental testing. Then, a series of data are collected from simulation to build the baseline and fault databases under a fixed load profile. Next, promising features are extracted from sensed parameters, and further data analysis are conducted to estimate the current health condition and to predict the remaining useful life of the target system. Some highlights of the work are included but not only limited to the following aspects: first, the methodology is based on electronic system simulation instead of traditional accelerated testing by employing a high-fidelity system simulation model and empirical critical component degradation models; second, efforts are made in this preliminary work to adapt proven prognostics and health management techniques from machinery to electronic health management, with the goal of expanding the realm of electronic diagnostics and prognostics.

Annual Conference of the PHM Society, 2010

Machine remaining useful life (RUL) prediction is a key part of Condition-Based Maintenance (CBM)... more Machine remaining useful life (RUL) prediction is a key part of Condition-Based Maintenance (CBM), which provides the time evolution of the fault indicator so that maintenance can be performed to avoid catastrophic failures. This paper proposes a new RUL prediction method based on adaptive neuro-fuzzy inference systems (ANFIS) and high-order particle filtering, which predicts the time evolution of the fault indicator and computes the probability density function (pdf) of RUL. The ANFIS is trained and integrated in a high-order particle filter to describe the fault propagation process; the high-order particle filter uses real-time data to update the current state estimates so as to improve the prediction accuracy. The performance of the proposed method is evaluated via the real-world data from a seeded fault test for a UH-60 helicopter planetary gear plate. The results show that it outperforms the conventional ANFIS predictor.

Annual Conference of the PHM Society, Oct 2, 2014

A direct method of measuring corrosion on a structure using a micro-linear polarization resistanc... more A direct method of measuring corrosion on a structure using a micro-linear polarization resistance (µLPR) sensor is presented. The new three-electrode µLPR sensor design presented in this paper improves on existing LPR sensor technology by using the structure as part of the sensor system, allowing the sensor electrodes to be made from a corrosion resistant or inert metal. This is in contrast to a twoelectrode µLPR sensor where the electrodes are made from the same material as the structure. A controlled experiment, conducted using an ASTM B117 salt fog, demonstrated the three-electrode µLPR sensors have a longer lifetime and better performance when compared to the two-electrode µLPR sensors. Following this evaluation, a controlled experiment using the ASTM G85 Annex 5 standard was performed to evaluate the accuracy and precision of the three-electrode µLPR sensor when placed between lap joint specimens made from AA7075-T6. The corrosion computed from the µLPR sensors agreed with the coupon mass loss to within a 95% confidence interval. Following the experiment, the surface morphology of each lap joint was determined using laser microscopy and stylus-based profilometry to obtain local and global surface images of the test panels. Image processing, feature extraction, and selection tools were then employed to identify the corrosion mechanism (e.g. pitting, intergranular). Douglas Brown 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.

Annual Conference of the Prognostics and Health Management Society, Oct 14, 2013

Most fault adaptive control research addresses the preservation of system stability or functional... more Most fault adaptive control research addresses the preservation of system stability or functionality in the presence of a specific failure (fault). This paper ex- amines the fault adaptive control problem for a generic class of incipient failure modes, which do not initially affect system stability, but will eventually cause a catas- trophic failure to occur. This risk of catastrophic failure due a component fault mode is some monotonically in- creasing function of the load on the component. Assum- ing that a probabilistic prognostic model is available to evaluate the risk of incipient fault modes growing into catastrophic failure conditions, then fundamentally the fault adaptive control problem is to adjust component loads to minimize risk of failure, while not overly de- grading nominal performance. A methodology is pro- posed for posing this problem as a finite horizon con- strained optimization, where constraints correspond to maximum risk of failure and maximum deviation from n...

International Journal of Prognostics and Health Management, 2020

It has been established that corrosion is one of the most important factors causing deterioration... more It has been established that corrosion is one of the most important factors causing deterioration and decreased performance and reliability in critical aerospace and industrial systems. Corrosion monitoring, detection, and quantification are recognized as key enabling technologies to reduce the impact of corrosion on the integrity of aircraft and industrial assets. Accurate and reliable detection of corrosion initiation and propagation, with specified false alarm rates, requires novel tools and methods, including verifiable simulation and modeling methods. This paper reports an experimental investigation of the detection and quantification of pitting corrosion on aluminum alloy panels using 3D surface metrology methods and image processing techniques. Panel surfaces were evaluated by laser microscopy and stylus-based profilometry to characterize global and local surface features. Promising imaging and texture features were extracted and compared between coated and uncoated aluminum ...

International Journal of Prognostics and Health Management, 2020

After an incipient fault mode has been detected a logical question to ask is: How long can the sy... more After an incipient fault mode has been detected a logical question to ask is: How long can the system continue to be operated before the incipient fault mode degrades to a failure condition? In many cases answering this question is complicated by the fact that further fault growth will depend on how the system is intended to be used in the future. The problem is then complicated even further when we consider that the future operation of a system may itself be conditioned on estimates of a system’s current health and on predictions of future fault evolution. This paper introduces a notationally convenient formulation of this problem as a Markov decision process. Prognostics-based fault management policies are then shown to be identified using standard Markov decision process optimization techniques. A case study example is analyzed, in which a discrete random walk is used to represent time-varying system loading demands. A comparison of fault management policies computed with and wit...

Studies in Informatics and Control, 2010

Failure prognosis, and particularly representation and management of uncertainty in long-term pre... more Failure prognosis, and particularly representation and management of uncertainty in long-term predictions, is a topic of paramount importance not only to improve productivity and efficiency, but also to ensure safety in the system's operation. The use of particle filter (PF) algorithms-in combination with outer feedback correction loopshas contributed significantly to the development of a robust framework for online estimation of the remaining useful equipment life. This paper explores the advantages and disadvantages of a Risk-Sensitive PF (RSPF) prognosis framework that complements the benefits of the classic approach, by representing the probability of rare events and highly non-monotonic phenomena within the formulation of the nonlinear dynamic equation that describes the evolution of the fault condition in time. The performance of this approach is thoroughly compared using a set of ad-hoc metrics. Actual data illustrating aging of an energy storage device (specifically battery capacity measurements [A-hr]) are used to test the proposed framework.

2015 IEEE AUTOTESTCON, 2015

This paper introduces a novel architecture for human-machine interface focusing primarily on the ... more This paper introduces a novel architecture for human-machine interface focusing primarily on the human aspects as applied to aircraft and unmanned systems. There is a need to explore new human-machine interface strategies stemming from the proliferation over the past years of accidents due to system complexity, failure modes and human errors. Concepts of autonomy establish the foundational elements of the work. We pursue a rigorous systems engineering process to analyze and design the tools and techniques for automated vehicle health monitoring, human-automation interface and conflict resolution enabled by innovative methods from game theory and reasoning algorithms. The general structure is illustrated in the paper. This paper addresses the general interface framework while emphasizing the human's (pilots) intended actions following an adverse event on-board the vehicle, i.e. critical component fault/failure modes. When combined with automated health state assessment means on-board the aircraft, the proposed strategy assists to improve the reliability of estimated actions the pilot must execute to mitigate possible catastrophic consequences. A “smart” knowledge base is exploited as the reasoning paradigm where cases are stored and new ones are compared with similar ones available in the case library. Learning and adaptation tools are used to improve the decision making process. The emphasis of this contribution is on methods and tools for conflict resolution when the automated system's advisories are coincident with the human's intended actions. Appropriate similarity metrics are defined and used for this purpose. The efficacy of the approach is demonstrated via an interface built in MATLAB highlighting the performance of the algorithmic modules for assessment and conflict resolution.

2008 International Conference on Prognostics and Health Management, 2008

IEEE Transactions on Biomedical Engineering, 2003

Expert Systems with Applications, 2013

In this research work a new hybrid approach to spatio-temporal seismic clustering is proposed. Th... more In this research work a new hybrid approach to spatio-temporal seismic clustering is proposed. The method builds upon a novel density based clustering scheme that explicitly takes into account earthquake's magnitude during the density estimation. The new density based clustering algorithm considers both time and spatial information during cluster formation. Therefore clusters lie in a spatio-temporal space. A hierarchical agglomerative clustering algorithm acts upon the identified clusters after dropping the time information in order to come up only with the spatial description of seismic events. The approach is demonstrated using data from the vicinity of the Hellenic seismic arc in order to enable its comparison with some of the state-of-the-art distinct seismic region identification methodologies. The presented results indicate that the combination of the two clustering stages could be potentially used for an automatic definition of major seismic sources.

Clinical Neurophysiology, 2005

Objective: To develop a prospective method for optimizing seizure prediction, given an array of i... more Objective: To develop a prospective method for optimizing seizure prediction, given an array of implanted electrodes and a set of candidate quantitative features computed at each contact location. Methods: The method employs a genetic-based selection process, and then tunes a probabilistic neural network classifier to predict seizures within a 10 min prediction horizon. Initial seizure and interictal data were used for training, and the remaining IEEG data were used for testing. The method continues to train and learn over time. Results: Validation of these results over two workshop patients demonstrated a sensitivity of 100%, and 1.1 false positives per hour for Patient E, using a 2.4 s block predictor, and a failure of the method on Patient B. Conclusions: This study demonstrates a prospective, exploratory implementation of a seizure prediction method designed to adapt to individual patients with a wide variety of pre-ictal patterns, implanted electrodes and seizure types. Its current performance is limited likely by the small number of input channels and quantitative features employed in this study, and segmentation of the data set into training and testing sets rather than using all continuous data available. Significance: This technique theoretically has the potential to address the challenge presented by the heterogeneity of EEG patterns seen in medication-resistant epilepsy. A more comprehensive implementation utilizing all electrode sites, a broader feature library, and automated multi-feature fusion will be required to fully judge the method's potential for predicting seizures.

Clinical Neurophysiology, 2005

Objective: Increases in accumulated energy on intracranial EEG are associated with oncoming seizu... more Objective: Increases in accumulated energy on intracranial EEG are associated with oncoming seizures in retrospective studies, supporting the idea that seizures are generated over time. Published seizure prediction methods require comparison to 'baseline' data, sleep staging, and selecting seizures that are not clustered closely in time. In this study, we attempt to remove these constraints by using a continuously adapting energy threshold, and to identify stereotyped energy variations through the seizure cycle (inter-, pre-, post-and ictal periods). Methods: Accumulated energy was approximated by using moving averages of signal energy, computed for window lengths of 1 and 20 min, and an adaptive decision threshold. Predictions occurred when energy within the shorter running window exceeded the decision threshold. Results: Predictions for time horizons of less than 3 h did not achieve statistical significance in the data sets analyzed that had an average inter-seizure interval ranging from 2.9 to 8.6 h. 51.6% of seizures across all patients exhibited stereotyped pre-ictal energy bursting and quiet periods. Conclusions: Accumulating energy alone is not sufficient for predicting seizures using a 20 min running baseline for comparison. Stereotyped energy patterns through the seizure cycle may provide clues to mechanisms underlying seizure generation. Significance: Energy-based seizure prediction will require fusion of multiple complimentary features and perhaps longer running averages to compensate for post-ictal and sleep-induced energy changes.

Fault diagnosis and failure prognosis of critical dynamic systems, such as aircraft and industria... more Fault diagnosis and failure prognosis of critical dynamic systems, such as aircraft and industrial processes, rely on degradation or fatigue models and measurements typically acquired on-line in real-time. Such sensor data must be pre-processed in order to remove artifacts and improve the signal-to-noise ratio. Furthermore, they must be processed appropriately so that useful information in compact form can be extracted and used to detect incipient failures and predict the remaining useful life of failing components. We present a methodology to select an optimum feature vector from a list of candidate features, prioritize and rank them to meet set performance objectives. The enabling technologies include genetic programming tools, data fusion and model-based approaches for feature selection and extraction. We will suggest a multi-core processing environment for the efficient and expedient implementation of these technologies. Performance metrics are defined to assess the efficacy of the methodology. Typical examples from aircraft systems are used to demonstrate the proposed techniques.

Proceedings of the ... American Control Conference, Jul 1, 2007

Critical aircraft assets are required to be available when needed, while exhibiting attributes of... more Critical aircraft assets are required to be available when needed, while exhibiting attributes of reliability, robustness, and high confidence under a variety of flight regimes, and maintained on the basis of their current condition rather than on the basis of scheduled maintenance practices. New and innovative technologies must be developed and implemented to address these concerns. Condition-based maintenance requires that the health of critical components/systems be monitored and diagnostic/prognostic strategies be developed to detect and identify incipient failures and predict the failing component's remaining useful life. Typically, vibration and other key indicators onboard an aircraft are severely corrupted by noise, thus curtailing the ability to accurately diagnose and predict failures. This paper introduces a novel blind deconvolution denoising scheme that employs a vibration model in the frequency domain and attempts to arrive at the true vibration signal through an iterative optimization process. Performance indexes are defined and data from a helicopter are used to demonstrate the effectiveness of the proposed approach. Index Terms-Blind deconvolution, planetary gear train, vibration signal denoising. I. INTRODUCTION F AULT DETECTION, isolation, and identification, especially for safety critical components and subsystems, have recently drawn increasing interest in the condition-based maintenance (CBM) community [1]-[4]. Epicyclic, or planetary, gear trains are widely used in the main transmission of many systems, such as the helicopter and other aircraft [5], [7]. This kind of gear system consists of an inner sun gear, two or more rotating planet gears, a stationary outer ring gear, and a planetary carrier. The planet gears, which surround the sun gear, are riding on the planetary carrier and also rotating inside the outer ring gear. In the operation, torque is transmitted through the sun gear to the planets and planetary carrier. The carrier, in turn, transmits torque to the main rotor shaft and blades [6]-[9]. The gearbox is a critical component that directly impacts the safety and performance of the aircraft.

IEEE Transactions on Biomedical Engineering, Aug 1, 2003

Springer eBooks, 2009

Mechanisms underlying seizure generation are traditionally thought to act over seconds to minutes... more Mechanisms underlying seizure generation are traditionally thought to act over seconds to minutes before clinical seizure onset. We analyzed continuous 3- to 14-day intracranial EEG recordings from five patients with mesial temporal lobe epilepsy obtained during evaluation for epilepsy surgery. We found localized quantitative EEG changes identifying prolonged bursts of complex epileptiform discharges that became more prevalent 7 hr before seizures and highly localized subclinical seizure-like activity that became more frequent 2 hr prior to seizure onset. Accumulated energy increased in the 50 min before seizure onset, compared to baseline. These observations, from a small number of patients, suggest that epileptic seizures may begin as a cascade of electrophysiological events that evolve over hours and that quantitative measures of preseizure electrical activity could possibly be used to predict seizures far in advance of clinical onset.

Annual Conference of the PHM Society, Oct 2, 2014

With electrical power supplies playing an important role in the operation of aircraft systems and... more With electrical power supplies playing an important role in the operation of aircraft systems and subsystems , flight and ground crews need health state awareness and prediction tools that accurately diagnose faults, predict failures, and project remaining life of these onboard power supplies. Among onboard power supplies, switch-mode power supplies are commonly used where their weight, size, and efficiency make them preferable to conventional transformer-based power supplies. In this paper, we present a framework of diagnostics and prognostics methodology based on an equivalent circuit system simulation model developed from a commercially available switch-mode power supply, and empirical component degradation models. In industrial applications, case-specified modifications can be made according to specific experimental or service conditions of different commercial products. First, the developed simulation model is validated through experimental testing. Then, a series of data are collected from simulation to build the baseline and fault databases under a fixed load profile. Next, promising features are extracted from sensed parameters, and further data analysis are conducted to estimate the current health condition and to predict the remaining useful life of the target system. Some highlights of the work are included but not only limited to the following aspects: first, the methodology is based on electronic system simulation instead of traditional accelerated testing by employing a high-fidelity system simulation model and empirical critical component degradation models; second, efforts are made in this preliminary work to adapt proven prognostics and health management techniques from machinery to electronic health management, with the goal of expanding the realm of electronic diagnostics and prognostics.

Annual Conference of the PHM Society, 2010

Machine remaining useful life (RUL) prediction is a key part of Condition-Based Maintenance (CBM)... more Machine remaining useful life (RUL) prediction is a key part of Condition-Based Maintenance (CBM), which provides the time evolution of the fault indicator so that maintenance can be performed to avoid catastrophic failures. This paper proposes a new RUL prediction method based on adaptive neuro-fuzzy inference systems (ANFIS) and high-order particle filtering, which predicts the time evolution of the fault indicator and computes the probability density function (pdf) of RUL. The ANFIS is trained and integrated in a high-order particle filter to describe the fault propagation process; the high-order particle filter uses real-time data to update the current state estimates so as to improve the prediction accuracy. The performance of the proposed method is evaluated via the real-world data from a seeded fault test for a UH-60 helicopter planetary gear plate. The results show that it outperforms the conventional ANFIS predictor.

Annual Conference of the PHM Society, Oct 2, 2014

A direct method of measuring corrosion on a structure using a micro-linear polarization resistanc... more A direct method of measuring corrosion on a structure using a micro-linear polarization resistance (µLPR) sensor is presented. The new three-electrode µLPR sensor design presented in this paper improves on existing LPR sensor technology by using the structure as part of the sensor system, allowing the sensor electrodes to be made from a corrosion resistant or inert metal. This is in contrast to a twoelectrode µLPR sensor where the electrodes are made from the same material as the structure. A controlled experiment, conducted using an ASTM B117 salt fog, demonstrated the three-electrode µLPR sensors have a longer lifetime and better performance when compared to the two-electrode µLPR sensors. Following this evaluation, a controlled experiment using the ASTM G85 Annex 5 standard was performed to evaluate the accuracy and precision of the three-electrode µLPR sensor when placed between lap joint specimens made from AA7075-T6. The corrosion computed from the µLPR sensors agreed with the coupon mass loss to within a 95% confidence interval. Following the experiment, the surface morphology of each lap joint was determined using laser microscopy and stylus-based profilometry to obtain local and global surface images of the test panels. Image processing, feature extraction, and selection tools were then employed to identify the corrosion mechanism (e.g. pitting, intergranular). Douglas Brown 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.

Annual Conference of the Prognostics and Health Management Society, Oct 14, 2013

Most fault adaptive control research addresses the preservation of system stability or functional... more Most fault adaptive control research addresses the preservation of system stability or functionality in the presence of a specific failure (fault). This paper ex- amines the fault adaptive control problem for a generic class of incipient failure modes, which do not initially affect system stability, but will eventually cause a catas- trophic failure to occur. This risk of catastrophic failure due a component fault mode is some monotonically in- creasing function of the load on the component. Assum- ing that a probabilistic prognostic model is available to evaluate the risk of incipient fault modes growing into catastrophic failure conditions, then fundamentally the fault adaptive control problem is to adjust component loads to minimize risk of failure, while not overly de- grading nominal performance. A methodology is pro- posed for posing this problem as a finite horizon con- strained optimization, where constraints correspond to maximum risk of failure and maximum deviation from n...

International Journal of Prognostics and Health Management, 2020

It has been established that corrosion is one of the most important factors causing deterioration... more It has been established that corrosion is one of the most important factors causing deterioration and decreased performance and reliability in critical aerospace and industrial systems. Corrosion monitoring, detection, and quantification are recognized as key enabling technologies to reduce the impact of corrosion on the integrity of aircraft and industrial assets. Accurate and reliable detection of corrosion initiation and propagation, with specified false alarm rates, requires novel tools and methods, including verifiable simulation and modeling methods. This paper reports an experimental investigation of the detection and quantification of pitting corrosion on aluminum alloy panels using 3D surface metrology methods and image processing techniques. Panel surfaces were evaluated by laser microscopy and stylus-based profilometry to characterize global and local surface features. Promising imaging and texture features were extracted and compared between coated and uncoated aluminum ...

International Journal of Prognostics and Health Management, 2020

After an incipient fault mode has been detected a logical question to ask is: How long can the sy... more After an incipient fault mode has been detected a logical question to ask is: How long can the system continue to be operated before the incipient fault mode degrades to a failure condition? In many cases answering this question is complicated by the fact that further fault growth will depend on how the system is intended to be used in the future. The problem is then complicated even further when we consider that the future operation of a system may itself be conditioned on estimates of a system’s current health and on predictions of future fault evolution. This paper introduces a notationally convenient formulation of this problem as a Markov decision process. Prognostics-based fault management policies are then shown to be identified using standard Markov decision process optimization techniques. A case study example is analyzed, in which a discrete random walk is used to represent time-varying system loading demands. A comparison of fault management policies computed with and wit...

Studies in Informatics and Control, 2010

Failure prognosis, and particularly representation and management of uncertainty in long-term pre... more Failure prognosis, and particularly representation and management of uncertainty in long-term predictions, is a topic of paramount importance not only to improve productivity and efficiency, but also to ensure safety in the system's operation. The use of particle filter (PF) algorithms-in combination with outer feedback correction loopshas contributed significantly to the development of a robust framework for online estimation of the remaining useful equipment life. This paper explores the advantages and disadvantages of a Risk-Sensitive PF (RSPF) prognosis framework that complements the benefits of the classic approach, by representing the probability of rare events and highly non-monotonic phenomena within the formulation of the nonlinear dynamic equation that describes the evolution of the fault condition in time. The performance of this approach is thoroughly compared using a set of ad-hoc metrics. Actual data illustrating aging of an energy storage device (specifically battery capacity measurements [A-hr]) are used to test the proposed framework.

2015 IEEE AUTOTESTCON, 2015

This paper introduces a novel architecture for human-machine interface focusing primarily on the ... more This paper introduces a novel architecture for human-machine interface focusing primarily on the human aspects as applied to aircraft and unmanned systems. There is a need to explore new human-machine interface strategies stemming from the proliferation over the past years of accidents due to system complexity, failure modes and human errors. Concepts of autonomy establish the foundational elements of the work. We pursue a rigorous systems engineering process to analyze and design the tools and techniques for automated vehicle health monitoring, human-automation interface and conflict resolution enabled by innovative methods from game theory and reasoning algorithms. The general structure is illustrated in the paper. This paper addresses the general interface framework while emphasizing the human's (pilots) intended actions following an adverse event on-board the vehicle, i.e. critical component fault/failure modes. When combined with automated health state assessment means on-board the aircraft, the proposed strategy assists to improve the reliability of estimated actions the pilot must execute to mitigate possible catastrophic consequences. A “smart” knowledge base is exploited as the reasoning paradigm where cases are stored and new ones are compared with similar ones available in the case library. Learning and adaptation tools are used to improve the decision making process. The emphasis of this contribution is on methods and tools for conflict resolution when the automated system's advisories are coincident with the human's intended actions. Appropriate similarity metrics are defined and used for this purpose. The efficacy of the approach is demonstrated via an interface built in MATLAB highlighting the performance of the algorithmic modules for assessment and conflict resolution.

2008 International Conference on Prognostics and Health Management, 2008

IEEE Transactions on Biomedical Engineering, 2003

Expert Systems with Applications, 2013

In this research work a new hybrid approach to spatio-temporal seismic clustering is proposed. Th... more In this research work a new hybrid approach to spatio-temporal seismic clustering is proposed. The method builds upon a novel density based clustering scheme that explicitly takes into account earthquake's magnitude during the density estimation. The new density based clustering algorithm considers both time and spatial information during cluster formation. Therefore clusters lie in a spatio-temporal space. A hierarchical agglomerative clustering algorithm acts upon the identified clusters after dropping the time information in order to come up only with the spatial description of seismic events. The approach is demonstrated using data from the vicinity of the Hellenic seismic arc in order to enable its comparison with some of the state-of-the-art distinct seismic region identification methodologies. The presented results indicate that the combination of the two clustering stages could be potentially used for an automatic definition of major seismic sources.

Clinical Neurophysiology, 2005

Objective: To develop a prospective method for optimizing seizure prediction, given an array of i... more Objective: To develop a prospective method for optimizing seizure prediction, given an array of implanted electrodes and a set of candidate quantitative features computed at each contact location. Methods: The method employs a genetic-based selection process, and then tunes a probabilistic neural network classifier to predict seizures within a 10 min prediction horizon. Initial seizure and interictal data were used for training, and the remaining IEEG data were used for testing. The method continues to train and learn over time. Results: Validation of these results over two workshop patients demonstrated a sensitivity of 100%, and 1.1 false positives per hour for Patient E, using a 2.4 s block predictor, and a failure of the method on Patient B. Conclusions: This study demonstrates a prospective, exploratory implementation of a seizure prediction method designed to adapt to individual patients with a wide variety of pre-ictal patterns, implanted electrodes and seizure types. Its current performance is limited likely by the small number of input channels and quantitative features employed in this study, and segmentation of the data set into training and testing sets rather than using all continuous data available. Significance: This technique theoretically has the potential to address the challenge presented by the heterogeneity of EEG patterns seen in medication-resistant epilepsy. A more comprehensive implementation utilizing all electrode sites, a broader feature library, and automated multi-feature fusion will be required to fully judge the method's potential for predicting seizures.

Clinical Neurophysiology, 2005

Objective: Increases in accumulated energy on intracranial EEG are associated with oncoming seizu... more Objective: Increases in accumulated energy on intracranial EEG are associated with oncoming seizures in retrospective studies, supporting the idea that seizures are generated over time. Published seizure prediction methods require comparison to 'baseline' data, sleep staging, and selecting seizures that are not clustered closely in time. In this study, we attempt to remove these constraints by using a continuously adapting energy threshold, and to identify stereotyped energy variations through the seizure cycle (inter-, pre-, post-and ictal periods). Methods: Accumulated energy was approximated by using moving averages of signal energy, computed for window lengths of 1 and 20 min, and an adaptive decision threshold. Predictions occurred when energy within the shorter running window exceeded the decision threshold. Results: Predictions for time horizons of less than 3 h did not achieve statistical significance in the data sets analyzed that had an average inter-seizure interval ranging from 2.9 to 8.6 h. 51.6% of seizures across all patients exhibited stereotyped pre-ictal energy bursting and quiet periods. Conclusions: Accumulating energy alone is not sufficient for predicting seizures using a 20 min running baseline for comparison. Stereotyped energy patterns through the seizure cycle may provide clues to mechanisms underlying seizure generation. Significance: Energy-based seizure prediction will require fusion of multiple complimentary features and perhaps longer running averages to compensate for post-ictal and sleep-induced energy changes.