Fault Diagnosis of a Two Spool Turbo-Fan Engine Using Transient Data: A Genetic Algorithm Approach

2003 ◽  
Author(s):  
Suresh Sampath ◽  
Y. G. Li ◽  
S. O. T. Ogaji ◽  
Riti Singh
Keyword(s):  
Genetic Algorithm ◽  
Fault Diagnosis ◽  
Simulated Data ◽  
Operating Conditions ◽  
Turbofan Engine ◽  
Transient Model ◽  
Turbine Performance ◽  
Genetic Algorithm Approach ◽  
Transient Data ◽  
Cumulative Deviation

Traditionally engine fault diagnosis has been performed at steady state conditions. There are several problems which can only be detected by transient data analysis like bearing fault, some control problems etc.. In addition, gas turbine performance deviation due to a component fault is more likely to be magnified during transients, when compared with the same parameter deviations at steady states. The specific approach used in this paper is to compare model-based information with measured data obtained from the engine during a slam acceleration. The measured transient data (from actual engine) is compared with a set of simulated data from the engine transient model, under similar operating conditions and known faults through a Cumulative Deviation. The Cumulative Deviations obtained from the comparisons are minimized for the best match using Genetic Algorithm. The Genetic Algorithm has been tailored to use real coding [1] method and to meet the requirements of the new procedure. The paper describes the application of the approach to a 2-spool turbofan engine and discusses the preliminary studies conducted.

A Comparison of Hybrid Approaches for Turbofan Engine Gas Path Fault Diagnosis

2016 ◽  
Vol 33 (3) ◽  
Author(s):  
Feng Lu ◽  
Yafan Wang ◽  
Jinquan Huang ◽  
Qihang Wang
Keyword(s):  
Genetic Algorithm ◽  
Kalman Filter ◽  
Fault Diagnosis ◽  
Fitness Function ◽  
Performance Degradation ◽  
Estimation Accuracy ◽  
Adaptive Genetic Algorithm ◽  
Turbofan Engine ◽  
Gas Path Fault ◽  
Fault Mode

AbstractA hybrid diagnostic method utilizing Extended Kalman Filter (EKF) and Adaptive Genetic Algorithm (AGA) is presented for performance degradation estimation and sensor anomaly detection of turbofan engine. The EKF is used to estimate engine component performance degradation for gas path fault diagnosis. The AGA is introduced in the integrated architecture and applied for sensor bias detection. The contributions of this work are the comparisons of Kalman Filters (KF)-AGA algorithms and Neural Networks (NN)-AGA algorithms with a unified framework for gas path fault diagnosis. The NN needs to be trained off-line with a large number of prior fault mode data. When new fault mode occurs, estimation accuracy by the NN evidently decreases. However, the application of the Linearized Kalman Filter (LKF) and EKF will not be restricted in such case. The crossover factor and the mutation factor are adapted to the fitness function at each generation in the AGA, and it consumes less time to search for the optimal sensor bias value compared to the Genetic Algorithm (GA). In a word, we conclude that the hybrid EKF-AGA algorithm is the best choice for gas path fault diagnosis of turbofan engine among the algorithms discussed.

Gas Path Fault Diagnosis of a Turbofan Engine From Transient Data Using Artificial Neural Networks

2003 ◽  
Author(s):  
S. O. T. Ogaji ◽  
Y. G. Li ◽  
S. Sampath ◽  
R. Singh
Keyword(s):  
Neural Networks ◽  
High Pressure ◽  
Artificial Neural Networks ◽  
Steady State ◽  
Fault Diagnosis ◽  
Turbofan Engine ◽  
Transient Data ◽  
Gas Path Fault ◽  
Pressure Compressor ◽  
Trained Network

Transient and steady state data may contain the same essential fault information but some faults have been shown to be more easily detectable from transient data because the transient records provide significant diagnostic content especially as the fault effects are magnified under transient. Various traditional and conventional techniques such as fault trees, fault matrixes, gas path analysis and its variants have been applied to gas path fault diagnosis of gas turbines. Recently, artificial intelligence techniques such as artificial neural networks (ANN) as well as optimization techniques such as genetic algorithm (GA) are being explored for fault diagnosis activities. In this paper, a novel approach to gas path fault diagnosis is proposed. The method involves the use of ANN with engine transient data. A set of nested neural networks designed to estimate independent parameter (efficiencies and flow capacities) changes due to faults within single or multiple components of a turbofan engine are presented. The approach involves classification and approximation type networks. Measurements from the engine are first assessed by a trained network and if a fault is diagnosed, are then classified into two groups — those originating from sensor faults and those from component faults, by another trained network. Other trained networks continue the fault isolation process and finally the magnitude of the fault(s) is quantified. A computer simulation of the process shows that results from a batched process of these networks can be obtained in less than three seconds. Four of the gas path components — intermediate pressure compressor (IPC), high pressure compressor (HPC), high pressure turbine (HPT) and low pressure turbine (LPT) — and measurements from eight sensors are considered. Sensor noise and bias are also considered in this analysis. The comparison of fault signatures from a steady state and transient process show that diagnosis with transient data can improve the accuracy of gas turbine fault diagnosis.

scholarly journals Data-Driven Approaches for Diagnosis of Incipient Faults in Cutting Arms of the Roadheader

2021 ◽  
Vol 2021 ◽  
pp. 1-15
Author(s):  
Qiang Liu ◽  
Songyong Liu ◽  
Qianjin Dai ◽  
Xiao Yu ◽  
Daoxiang Teng ◽  
...  
Keyword(s):  
Genetic Algorithm ◽  
Fault Diagnosis ◽  
Support Vector Machines ◽  
Back Propagation ◽  
Operating Conditions ◽  
Support Vector ◽  
Algorithm Optimization ◽  
Genetic Algorithm Optimization ◽  
Fault Detection And Identification ◽  
Vector Machines

Incipient fault detection and identification (IFDI) of cutting arms is a crucial guarantee for the smooth operation of a roadheader. However, the shortage of fault samples restricts the application of the fault diagnosis technique, and the data analysis tools should be optimized efficiently. In this study, four machine learning tools (the back-propagation neural network based on genetic algorithm optimization, the naive Bayes based on genetic algorithm optimization, the support vector machines based on particle swarm optimization, and the support vector machines based on dynamic cuckoo) are applied to address the challenge in the IFDI of cutting arms. The commonly measured current and vibration data cutting arms are used in the IFDI. The experimental results show that the support vector machines based on dynamic cuckoo outperform the other methods. Besides, the performance of the four methods under different operating conditions is compared. The fault cause of cutting arms of the roadheader is analyzed and the design improvement scheme for cutting arms is provided. This study provides a reference for improving the fault diagnosis of the roadheader.

Nonlinear Multiple Points Gas Turbine Off-Design Performance Adaptation Using a Genetic Algorithm

2011 ◽  
Vol 133 (7) ◽  
Author(s):  
Y. G. Li ◽  
M. F. Abdul Ghafir ◽  
L. Wang ◽  
R. Singh ◽  
K. Huang ◽  
...  
Keyword(s):  
Genetic Algorithm ◽  
Gas Turbine ◽  
Scaling Factor ◽  
Performance Model ◽  
Operating Conditions ◽  
Performance Models ◽  
Design Performance ◽  
Turbine Performance ◽  
Thermodynamic Performance ◽  
Performance Adaptation

Accurate gas turbine performance models are crucial in many gas turbine performance analysis and gas path diagnostic applications. With current thermodynamic performance modeling techniques, the accuracy of gas turbine performance models at off-design conditions is determined by engine component characteristic maps obtained in rig tests and these maps may not be available to gas turbine users or may not be accurate for individual engines. In this paper, a nonlinear multiple point performance adaptation approach using a genetic algorithm is introduced with the aim to improve the performance prediction accuracy of gas turbine engines at different off-design conditions by calibrating the engine performance models against available test data. Such calibration is carried out with introduced nonlinear map scaling factor functions by “modifying” initially implemented component characteristic maps in the gas turbine thermodynamic performance models. A genetic algorithm is used to search for an optimal set of nonlinear scaling factor functions for the maps via an objective function that measures the difference between the simulated and actual gas path measurements. The developed off-design performance adaptation approach has been applied to a model single spool turbo-shaft aero gas turbine engine and has demonstrated a significant improvement in the performance model accuracy at off-design operating conditions.

A Novel Approach to Mechanism Reduction Optimization for an Aviation Fuel/Air Reaction Mechanism Using a Genetic Algorithm

2004 ◽  
Vol 128 (2) ◽  
pp. 255-263 ◽  
Author(s):  
L. Elliott ◽  
D. B. Ingham ◽  
A. G. Kyne ◽  
N. S. Mera ◽  
M. Pourkashanian ◽  
...  
Keyword(s):  
Genetic Algorithm ◽  
Reaction Mechanism ◽  
Reaction Rate ◽  
Flame Structure ◽  
Chemical Kinetic ◽  
Reaction Scheme ◽  
Operating Conditions ◽  
Aviation Fuel ◽  
Reduced Mechanism ◽  
Genetic Algorithm Approach

This study presents a novel multiobjective genetic-algorithm approach to produce a new reduced chemical kinetic reaction mechanism to simulate aviation fuel combustion under various operating conditions. The mechanism is used to predict the flame structure of an aviation fuel/O2∕N2 flame in both spatially homogeneous and one-dimensional premixed combustion. Complex hydrocarbon fuels, such as aviation fuel, involve large numbers of reaction steps with many species. As all the reaction rate data are not well known, there is a high degree of uncertainty in the results obtained using these large detailed reaction mechanisms. In this study a genetic algorithm approach is employed for determining new reaction rate parameters for a reduced reaction mechanism for the combustion of aviation fuel-air mixtures. The genetic algorithm employed incorporates both perfectly stirred reactor and laminar premixed flame data in the inversion process, thus producing an efficient reaction mechanism. This study provides an optimized reduced aviation fuel-air reaction scheme whose performance in predicting experimental major species profiles and ignition delay times is not only an improvement on the starting reduced mechanism but also on the full mechanism.

Non-Linear Multiple Points Gas Turbine Off-Design Performance Adaptation Using a Genetic Algorithm

2010 ◽  
Author(s):  
Y. G. Li ◽  
M. F. Abdul Ghafir ◽  
L. Wang ◽  
R. Singh ◽  
K. Huang ◽  
...  
Keyword(s):  
Genetic Algorithm ◽  
Gas Turbine ◽  
Scaling Factor ◽  
Operating Conditions ◽  
Performance Models ◽  
Design Performance ◽  
Turbine Performance ◽  
Thermodynamic Performance ◽  
Non Linear ◽  
Performance Adaptation

Accurate gas turbine performance models are crucial in many gas turbine performance analysis and gas path diagnostic applications. With current thermodynamic performance modelling techniques, the accuracy of gas turbine performance models at off-design conditions is determined by engine component characteristic maps obtained in rig tests and these maps may not be available to gas turbine users or may not be accurate for individual engines. In this paper, a non-linear multiple point performance adaptation approach using a Genetic Algorithm is introduced with the aim to improve the performance prediction accuracy of gas turbine engines at different off-design conditions by calibrating the engine performance models against available test data. Such calibration is carried out with introduced non-linear map scaling factor functions by ‘modifying’ initially implemented component characteristic maps in the gas turbine thermodynamic performance models. A Genetic Algorithm is used to search for an optimal set of non-linear scaling factor functions for the maps via an objective function that measures the difference between the simulated and actual gas path measurements. The developed off-design performance adaptation approach has been applied to a model single spool turboshaft aero gas turbine engine and demonstrated a significant improvement in the performance model accuracy at off-design operating conditions.

scholarly journals A Genetic Algorithm Optimized RNN-LSTM Model for Remaining Useful Life Prediction of Turbofan Engine

Electronics ◽  
2021 ◽  
Vol 10 (3) ◽  
pp. 285
Author(s):  
Kwok Tai Chui ◽  
Brij B. Gupta ◽  
Pandian Vasant
Keyword(s):  
Genetic Algorithm ◽  
Feature Extraction ◽  
Prediction Model ◽  
Prediction Models ◽  
Remaining Useful Life ◽  
Prediction Algorithm ◽  
Short Term ◽  
Turbofan Engine ◽  
Term Prediction ◽  
Useful Life

Understanding the remaining useful life (RUL) of equipment is crucial for optimal predictive maintenance (PdM). This addresses the issues of equipment downtime and unnecessary maintenance checks in run-to-failure maintenance and preventive maintenance. Both feature extraction and prediction algorithm have played crucial roles on the performance of RUL prediction models. A benchmark dataset, namely Turbofan Engine Degradation Simulation Dataset, was selected for performance analysis and evaluation. The proposal of the combination of complete ensemble empirical mode decomposition and wavelet packet transform for feature extraction could reduce the average root-mean-square error (RMSE) by 5.14–27.15% compared with six approaches. When it comes to the prediction algorithm, the results of the RUL prediction model could be that the equipment needs to be repaired or replaced within a shorter or a longer period of time. Incorporating this characteristic could enhance the performance of the RUL prediction model. In this paper, we have proposed the RUL prediction algorithm in combination with recurrent neural network (RNN) and long short-term memory (LSTM). The former takes the advantages of short-term prediction whereas the latter manages better in long-term prediction. The weights to combine RNN and LSTM were designed by non-dominated sorting genetic algorithm II (NSGA-II). It achieved average RMSE of 17.2. It improved the RMSE by 6.07–14.72% compared with baseline models, stand-alone RNN, and stand-alone LSTM. Compared with existing works, the RMSE improvement by proposed work is 12.95–39.32%.

Sign in / Sign up

Export Citation Format

Share Document