A Generalized Fault Classification for Gas Turbine Diagnostics on Steady States and Transients

2006 ◽  
Author(s):  
Igor Loboda ◽  
Sergey Yepifanov ◽  
Yakov Feldshteyn
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Steady State ◽  
Gas Turbine ◽  
Correct Diagnosis ◽  
Diagnostic Techniques ◽  
Operating Conditions ◽  
Fault Classification ◽  
Diagnostic Process ◽  
Transient Operation

Gas turbine diagnostic techniques are often based on the recognition methods using the deviations between actual and expected thermodynamic performances. The problem is that the deviations depend on real operating conditions. However, our studies show that such a dependency can be reduced. In this paper, we propose the generalized fault classification that is independent of the operating conditions. To prove this idea, the averaged probabilities of the correct diagnosis are computed and compared for two cases: the proposed classification and the traditional one based on the fixed operating conditions. The probabilities are calculated through a stochastic modeling of the diagnostic process, in which a thermodynamic model generates deviations that are induced by the faults. Artificial neural networks recognize these faults. The proposed classification principle has been realized for both, steady state and transient operation of the gas turbine units. The results show that the acceptance of the generalized classification practically does not reduce the diagnosis trustworthiness.

A Generalized Fault Classification for Gas Turbine Diagnostics at Steady States and Transients

2007 ◽  
Vol 129 (4) ◽  
pp. 977-985 ◽  
Author(s):  
Igor Loboda ◽  
Sergiy Yepifanov ◽  
Yakov Feldshteyn
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Steady State ◽  
Gas Turbine ◽  
Diagnostic Techniques ◽  
Fault Classification ◽  
Diagnostic Process ◽  
Practical Realization ◽  
Transient Operation ◽  
Operational Conditions

Gas turbine diagnostic techniques are often based on the recognition methods using the deviations between actual and expected thermodynamic performances. The problem is that the deviations generally depend on current operational conditions. However, our studies show that such a dependency can be low. In this paper, we propose a generalized fault classification that is independent of the operational conditions. To prove this idea, the probabilities of true diagnosis were computed and compared for two cases: the proposed classification and the conventional one based on a fixed operating point. The probabilities were calculated through a stochastic modeling of the diagnostic process. In this process, a thermodynamic model generates deviations that are induced by the faults, and an artificial neural network recognizes these faults. The proposed classification principle has been implemented for both steady state and transient operation of the analyzed gas turbine. The results show that the adoption of the generalized classification hardly affects diagnosis trustworthiness and the classification can be proposed for practical realization.

An Estimation of 3-Way Catalyst Performance Using Artificial Neural Networks During Idle Speed

2004 ◽  
Author(s):  
P. N. Botsaris ◽  
D. Bechrakis ◽  
P. D. Sparis
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Steady State ◽  
Expert Knowledge ◽  
Operating Conditions ◽  
Experimental Results ◽  
Outlet Temperature ◽  
Practical Applications ◽  
Idle Speed ◽  
Artificial Neural

The intelligent control as fuzzy or artificial is based on either expert knowledge or experimental data and therefore it possesses intrinsic qualities like robustness and ease implementation. Lately, many researchers present studies aim to show that this kind of control can be used in practical applications such as the idle speed control problem in automotive industry. In this study, an estimation of an automobile three-way catalyst performance with artificial neural networks is presented. It may be an alternative approach for an on board diagnostic system (OBD) to predict the catalyst performance. This method was tested using data sets from two kind of catalysts, a brand new and an old one on a laboratory bench at idle speed. The catalyst operation during the “steady state” phase (the phase that the catalyst has reached its operating conditions and works normally) is examined. Further experiments are needed for different catalyst typed before the methods is proposed generally. It consists of 855 elements of catalyst inlet-outlet temperature difference (DT), hydrocarbons (HC), and carbon monoxide (CO) and carbon dioxide (CO2) emissions. The simulation: detects the values of HC, CO, CO2 using the DT as an input to our network forms a neural network. Results showed serious indications that artificial neural networks (or fuzzy logic control laws) could estimate the catalyst performance adequately depending their training process, if certain information about the catalyst system and the inputs and output of such system are known. In this study the “steady state” period experimental results are presented. In this paper the “steady state” period experimental results are presented.

Gas Turbine Diagnostics Under Variable Operating Conditions

2007 ◽  
Author(s):  
Igor Loboda ◽  
Yakov Feldshteyn ◽  
Sergiy Yepifanov
Keyword(s):  
Gas Turbine ◽  
Health Monitoring ◽  
Correct Diagnosis ◽  
Diagnostic System ◽  
Diagnostic Technique ◽  
Operating Conditions ◽  
Diagnostic Process ◽  
Ambient Conditions ◽  
Operating Points ◽  
Variable Operating Conditions

Operating conditions (control variables and ambient conditions) of gas turbine plants and engines vary considerably. The fact that health monitoring has to be uninterrupted creates the need for a run time diagnostic system to operate under any conditions. The diagnostic technique described in this paper utilizes the thermodynamic models in order to simulate gaspath faults and uses neural networks for the faults localization. This technique is repeatedly executed and the diagnoses are registered. On the basis of these diagnoses and beforehand known faults, the correct diagnosis probabilities are then calculated. The present paper analyses the influence of the operating conditions on a diagnostic process. In the technique, different options are simulated of a diagnostic treatment of the measured values obtained under variable operating conditions. The mentioned above probabilities help to compare these options. The main focus of the paper is on the so called multipoint (multimode) diagnosis that groups the data from different operating points (modes) to set only a single diagnosis.

Performance Estimation of Direct Methanol Fuel Cell Using Artificial Neural Network

2015 ◽  
Author(s):  
M. A. Rafe Biswas ◽  
Melvin D. Robinson
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Networks ◽  
Fuel Cell ◽  
Direct Methanol Fuel Cell ◽  
Cell Voltage ◽  
Operating Conditions ◽  
Direct Methanol ◽  
Artificial Neural ◽  
Methanol Fuel Cell

A direct methanol fuel cell can convert chemical energy in the form of a liquid fuel into electrical energy to power devices, while simultaneously operating at low temperatures and producing virtually no greenhouse gases. Since the direct methanol fuel cell performance characteristics are inherently nonlinear and complex, it can be postulated that artificial neural networks represent a marked improvement in performance prediction capabilities. Artificial neural networks have long been used as a tool in predictive modeling. In this work, an artificial neural network is employed to predict the performance of a direct methanol fuel cell under various operating conditions. This work on the experimental analysis of a uniquely designed fuel cell and the computational modeling of a unique algorithm has not been found in prior literature outside of the authors and their affiliations. The fuel cell input variables for the performance analysis consist not only of the methanol concentration, fuel cell temperature, and current density, but also the number of cells and anode flow rate. The addition of the two typically unconventional variables allows for a more distinctive model when compared to prior neural network models. The key performance indicator of our neural network model is the cell voltage, which is an average voltage across the stack and ranges from 0 to 0:8V. Experimental studies were carried out using DMFC stacks custom-fabricated, with a membrane electrode assembly consisting of an additional unique liquid barrier layer to minimize water loss through the cathode side to the atmosphere. To determine the best fit of the model to the experimental cell voltage data, the model is trained using two different second order training algorithms: OWO-Newton and Levenberg-Marquardt (LM). The OWO-Newton algorithm has a topology that is slightly different from the topology of the LM algorithm by the employment of bypass weights. It can be concluded that the application of artificial neural networks can rapidly construct a predictive model of the cell voltage for a wide range of operating conditions with an accuracy of 10−3 to 10−4. The results were comparable with existing literature. The added dimensionality of the number of cells provided insight into scalability where the coefficient of the determination of the results for the two multi-cell stacks using LM algorithm were up to 0:9998. The model was also evaluated with empirical data of a single-cell stack.

Fault Classification with the AID of Artificial Neural Networks

1991 ◽  
Vol 24 (6) ◽  
pp. 541-545
Author(s):  
D.M. Himmelblau ◽  
R.W. Barker ◽  
W. Suewatanakul
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Fault Classification ◽  
Artificial Neural

scholarly journals Fault Signatures Obtained From Fault Implant Tests on an F404 Engine

1992 ◽  
Author(s):  
Richard W. Eustace ◽  
Bruce A. Woodyatt ◽  
Graeme L. Merrington ◽  
Tony A. Runacres
Keyword(s):  
Steady State ◽  
Diagnostic Techniques ◽  
Turbine Engines ◽  
Diagnostic Process ◽  
Gas Turbine Engines ◽  
Implant Tests ◽  
Increased Sensitivity ◽  
Transient Data ◽  
Fault Diagnostic ◽  
Fault Information

The fault diagnostic process for gas turbine engines can be improved if data acquired by an on-board engine monitoring system (EMS) are utilised effectively. In the commercial transport field, techniques are available to extract engine condition assessment information from steady-state EMS data. In a military environment, steady-state data are not always available, and therefore it is desirable to extract at least some of the information from transient data, such as during take-off. Fault signatures are presented for an F404 engine based on fault implant tests in a sea-level-static (SLS) test-cell. A comparison is then made between the fault coverage capabilities of fault diagnostic techniques based on the use of steady-state engine data with those using transient data. The important conclusions to emerge from this work are that for the range of faults examined, not only is there similar fault information contained within the transient data but the faults can be detected with increased sensitivity using these data.

Sign in / Sign up

Export Citation Format

Share Document