Gas Turbine Fault Diagnosis Research Articles

Feature-level SMOTE: Augmenting fault samples in learnable feature space for imbalanced fault diagnosis of gas turbines

A challenge in gas turbine fault diagnosis is that labeled fault samples are relatively rare and much fewer than normal samples. Conventional data augmentation techniques generate fault samples in original data spaces, resulting in the issue that synthetic fault samples highly overlap with normal samples. Aiming at the issue, a feature-level data augmentation method, namely feature-level SMOTE, is developed by integrating deep Siamese multi-head self-attention network (DSMHSA) with synthetic minority over-sampling technique (SMOTE) to reduce inter-class imbalance and overlap simultaneously. First, the DSMHSA maps original data into a feature space with better inter-class separability, in which inter-class samples stay far away from one another. Second, the SMOTE generates synthetic fault samples in the well-separable space, in order to balance the data set. Finally, the effectiveness of the developed feature-level SMOTE in imbalanced fault diagnosis has been evaluated through two case studies including the real gas turbine fault dataset and the public robot execution failures dataset. To be specific, its average balanced accuracy is 90.38% on the gas turbine dataset, yielding 9.67%, 13.94%, and 12.39% improvements compared to the OUPS, A-SUWO, and NRAS, respectively.

An improved performance diagnostic method for industrial gas turbines with consideration of intake and exhaust system

Industrial gas turbines, as core equipment for efficient and clean utilization of energy, have been widespreadly applied in gas-steam combined cycle power plants. To ensure safe, stable, green and efficient operation, it is essential to establish a reliable industrial gas turbine fault diagnosis system. Industrial gas turbines need more flexible operation, due to requirements for power plant frequent frequency regulation and peak regulation in grid-supported mode. It is difficult to filter out quasi-steady-state operation data segment from collected operating data of industrial gas turbine, which makes the applicable boundary of traditional diagnostic methods narrower. For the actual problems to be solved urgently during diagnosis of industrial gas turbines, this paper proposes an improved performance diagnostic method for industrial gas turbines. First, a high-precision thermodynamic modeling method for industrial gas turbines was proposed with consideration of the effect of variable geometry compressor, and the established gas turbine thermodynamic model is validated by actual industrial gas turbine operation data. Next, an improved performance diagnostic method is established, and the health parameters for intake system and exhaust system is proposed in fault diagnosis for the first time and compressor variable geometry is also considered in the diagnostic process. Finally, through actual industrial gas turbine operation test, the effectiveness and reliability of the proposed diagnostic method to diagnose component performance degradation and component abrupt performance damage are verified. The proposed diagnostic method realizes real-time online quantitative diagnosis of all gas-path components under full operating conditions from the minimum IGV opening load to the IGV fully open base load.

A time-series turbofan engine successive fault diagnosis under both steady-state and dynamic conditions

In recent years there has been a growing interest in gas turbine fault diagnosis, especially under dynamic conditions, due to the evolving operating profile of gas turbines and the need to deploy computationally efficient and high-precision diagnostic solutions in real-time. One of the main challenges of fault diagnosis in real-time is the power imbalance between the compressor and turbine that occurs during transient operation. In addition, the heat soakage phenomenon characterizing the transient conditions has a substantial impact on the accuracy of the diagnosis. Finally, any sudden failure that might happen during transient operating conditions creates an additional challenge to fault diagnostics. The present study proposes a gas turbine diagnostic approach based on time-series measurements encapsulating steady-state and transient operating conditions. Specifically, the introduced novel approach is capable of quantifying the surplus/deficit of the power between the compressor and the turbine by utilizing the time-series data representing the observed deviations in the shaft rotational speed in order to determine the power balance in the shaft. The maximum diagnostic errors for constant fault and sudden failure are less than 0.006% during the dynamic maneuver. The results demonstrate and illustrate that the proposed method could effectively and accurately diagnose the severity of aero-engine faults at both steady-state and transient conditions. Therefore, this study has great potential for gas turbine practitioners since the diagnosis under transient conditions in real-time can enhance the capability of engine online condition monitoring and improve the condition-based maintenance of gas turbine assets.

A comparative study on class-imbalanced gas turbine fault diagnosis

Gas turbines are widely used in various fields, and the failure of gas turbines can cause catastrophic consequences. Health condition monitoring and fault diagnosis of gas turbines can detect faults timely, avoid serious faults, and significantly reduce maintenance costs. Thus, fault diagnosis of gas turbines has great significance. Current researches on gas turbine fault diagnosis mainly focus on the case of abundant fault samples. However, fault data are very rare and the number of normal samples is much larger than the number of fault samples in the industrial scene. This class-imbalance problem widely exists but is hardly focused in the field of gas turbine fault diagnosis. Aiming to solve this problem, this paper introduces the concept of class-imbalanced learning from the machine learning field, summarizes three kinds of class-imbalance addressment methods including oversampling, undersampling, and sample weighting, and proposes a new combination method of focal loss and random oversampling for addressing class-imbalance in deep neural networks, and performs a systematic comparative study on class-imbalanced gas turbine fault diagnosis. Experimental results show that class-imbalance can seriously reduce the fault diagnosis accuracy. Through these class-imbalance addressment methods, diagnosis accuracy is greatly improved. Comparative experiments also show that the proposed combination method can obtain the best diagnosis accuracy among all the compared methods in class-imbalanced situation. Through this comparative study, a detailed guideline for improving diagnosis accuracy under class-imbalanced circumstance is provided.

Gas turbine gas‐path fault diagnosis in power plant under transient operating condition with variable geometry compressor

AbstractTo extend useful life, improve availability, and reduce gas turbine maintenance costs, scholars have proposed many gas‐path fault diagnostic approaches on the basis of steady‐state operating conditions. However, gas turbines increasingly have to operate more flexibly in grid‐supported modes. In transient operating mode, the service life of the gas turbine will be consumed faster compared with that in base‐load operating mode. In addition, the use of a variable geometry compressor has introduced new uncertainty in the actual diagnostic process, making the applicable boundaries of the original gas‐path diagnosis methods narrower and narrower. Regarding the issue above, a novel approach is proposed for fault diagnosis of gas turbine gas‐path components in power plants under a transient operating condition with the variable geometry compressor for the first time. First, the mathematical relationship of the influence of the compressor inlet guide vane position on the compressor flow and efficiency characteristics is deduced. And a high‐precision thermodynamic model for the purpose of performance analysis and gas‐path fault diagnosis is established, which comprehensively considers the above interference factors and includes health parameters of the main gas‐path components. Second, the thermodynamic model‐based diagnosis strategy under transient operating conditions with the variable geometry compressor is proposed and component health parameters can be quantitatively obtained. At last, the effectiveness and reliability of the method proposed are validated through both simulation test and actual operation test with good accuracy and real‐time performance to diagnose single and multiple component performance degradations, and component abrupt fault.

Research on fault diagnosis of gas turbine rotor based on adversarial discriminative domain adaption transfer learning

In the process of gas turbine rotor fault diagnosis based on data-driven, transfer learning is an effective method to solve the lack of gas turbines labeled data, which will result in domain shifts due to the data distribution difference between source domain data and target domain data under variable working condition. A gas turbine fault diagnosis method based on Adversarial Discriminative Domain Adaptation Transfer Learning Network (ADDATLN) is put forward to reduce domain offsets and improve the gas turbine fault diagnosis accuracy. In the proposed method, pre-trained deep Convolutional Neural Networks (CNN) models in the source domain is transferred to target domain data, then deep adversarial training between the source domain and target domain is adopted to adaptively optimize the model parameters of the target domain network, with the purpose of reducing domain offsets and improving gas turbine fault classification accuracy. Field test experiment results on gas turbine rotor fault diagnosis under different working conditions show that the average accuracy of the proposed method reaches 96.45%, and the average accuracy of fault diagnosis on different gas turbines with the same type achieved 95.13%. The field test results confirm that the method effectively reduces the domain differences caused by varying working conditions and different gas turbines, and improves the accuracy of gas turbine rotor fault diagnosis under variable working condition and for different gas turbines with small samples.

Research on Improved Deep Convolutional Generative Adversarial Networks for Insufficient Samples of Gas Turbine Rotor System Fault Diagnosis

In gas turbine rotor systems, an intelligent data-driven fault diagnosis method is an important means to monitor the health status of the gas turbine, and it is necessary to obtain sufficient fault data to train the intelligent diagnosis model. In the actual operation of a gas turbine, the collected gas turbine fault data are limited, and the small and imbalanced fault samples seriously affect the accuracy of the fault diagnosis method. Focusing on the imbalance of gas turbine fault data, an Improved Deep Convolutional Generative Adversarial Network (Improved DCGAN) suitable for gas turbine signals is proposed here, and a structural optimization of the generator and a gradient penalty improvement in the loss function are introduced to generate effective fault data and improve the classification accuracy. The experimental results of the gas turbine test bench demonstrate that the proposed method can generate effective fault samples as a supplementary set of fault samples to balance the dataset, effectively improve the fault classification and diagnosis performance of gas turbine rotors in the case of small samples, and provide an effective method for gas turbine fault diagnosis.

Gas path component fault diagnosis of an industrial gas turbine under different load condition using online sequential extreme learning machine

One of the challenges that data-based gas path component monitoring systems are facing, is the poor performance in those operational conditions that are not considered in the training process. An approach could be retraining the model repeatedly with a large number of previous and fresh data from the scratch, which would require a lot of time and storage memory. To address this problem, a component fault diagnostic system based on the bank of online sequential extreme learning machines (OSELMs) are applied that can incrementally update with any number of new training samples. This system is adapted for a two-shaft industrial gas turbine that is used for power generation. Besides, an optimal set of measurements are selected as the input of the system using a hybrid approach based on the variable length genetic algorithm and extreme learning machine (VLGA-ELM). The performance of OSELM based diagnostic system is compared with several batch learning algorithms by analyzing the confusion matrix of all systems. The results show that the OSELM based approach reflects a promising compromise between the overall accuracy, false positive rate, F1-score and false negative rate, while having the least training time compared with the other systems.

Vibration fault signal analysis and diagnosis of flue gas turbine

Flue gas turbine is a kind of typical rotating machinery. It uses the flue gas expansion which contains a lot of heat energy and pressure energy generated in the catalytic cracking process to drive the axial-flow compressor or generator to rotate. Due to the complex structure, bad operating environment, poor stability, and large vibration of the flue gas turbine, its failure rate is high, which causes immeasurable economic losses and safety hazards. Therefore, how to effectively obtain, extract and identify the fault information of flue gas turbine is a difficult problem in the field of fault diagnosis technology. In this paper, aiming at two kinds of faults in the tp9-90 flue gas turbine system, i.e., rotor misalignment and unbalance causing excessive vibration, the corresponding geometric model and mathematical model are established, and the fault mechanism is revealed. Through on-line monitoring of flue gas turbine rotor vibration signal, the running state of flue gas turbine is comprehensively analyzed, and the technological transformation is carried out according to the fault causes, the improved flue gas machine runs smoothly, and the vibration and failure rate are obviously reduced. This study provides an important reference for condition monitoring and fault diagnosis of flue gas turbines.

A novel approach to gas turbine fault diagnosis based on learning of fault characteristic maps using hybrid residual compensation extreme learning machine-growing neural gas model

A novel approach to gas turbine fault diagnosis based on learning of fault characteristic maps using hybrid residual compensation extreme learning machine-growing neural gas model

A New Feature Selection-Aided Observer for Sensor Fault Diagnosis of an Industrial Gas Turbine

In complex systems, observers play a vital role in fault diagnosis and performance monitoring. Hence, the accuracy of these observers is very important. In this paper, an alternative solution to improve the observer accuracy is investigated. For this purpose, a feature selection-aided observer is proposed. This observer utilizes a feature-selection scheme that augments the sensor data with its time-domain features to observe the system. A decision tree algorithm selects the features needed for the observer, which act as soft sensors. In this scheme, a sliding mode observer is integrated with the decision tree method to select features for attaining more information from sensors (extracting two outputs from each sensor) which causes to increase the number of the observer equations. To implement this method, an augmented model for providing the statistical information is defined. To assess the performance of the introduced algorithm, an industrial twin shaft gas turbine is considered. In this model, the big data of the SGT 600 gas turbine is broken down into numerous data sets with a predefined sample time, then the subspace algorithm (N4SID) is utilized to obtain a state-space model by mean and variance of the inputs and outputs. Moreover, the algorithm is used to detect and diagnose sensors fault with a novel decision logic. Simulation results illustrate that the proposed observer is accurate and reliable for monitoring performance and diagnosing faults and failures.

Long-short term memory and gas path analysis based gas turbine fault diagnosis and prognosis

At present, the main purpose of gas turbine fault prediction is to predict the performance decline trend of the whole system, but the quantitative and thorough performance health index (PHI) research of every major component is lacking. Regarding the issue above, a long-short term memory and gas path analysis (GPA) based gas turbine fault diagnosis and prognosis method is proposed, which realizes the coupling of fault diagnosis and prognosis process. The measurable gas path parameters (GPPs) and the health parameters (HP) of every main component of the goal engine are obtained through the adaptive modeling strategy and the gas path diagnosis method based on the thermodynamic model. The predictive model of the Long-Short Term Memory (LSTM) network combines the measurable GPPs and the diagnostic HPs to predict the HPs of each major component in the future. Simulation experiments show that the proposed method can effectively diagnose and predict detailed, quantified, and accurate PHIs of the main components. Among them, the maximum root mean square error (RMSE) of the diagnosed component HPs do not exceed 0.193%. The RMSE of the best prediction model is 0.232%, 0.029%, 0.069%, and 0.043% in the HP prediction results of each component, respectively.

Development of L1-norm sliding mode observer for sensor fault diagnosis of an industrial gas turbine

In this article, a novel approach for model-based sensor fault detection and estimation of gas turbine is presented. The proposed method includes driving a state-space model of gas turbine, designing a novel L1-norm Lyapunov-based observer, and a decision logic which is based on bank of observers. The novel observer is designed using multiple Lyapunov functions based on L1-norm, reducing the estimation noise while increasing the accuracy. The L1-norm observer is similar to sliding mode observer in switching time. The proposed observer also acts as a low-pass filter, subsequently reducing estimation chattering. Since a bank of observers is required in model-based sensor fault detection, a bank of L1-norm observers is designed in this article. Corresponding to the use of the bank of observers, a two-step fault detection decision logic is developed. Furthermore, the proposed state-space model is a hybrid data-driven model which is divided into two models for steady-state and transient conditions, according to the nature of the gas turbine. The model is developed by applying a subspace algorithm to the real field data of SGT-600 (an industrial gas turbine). The proposed model was validated by applying to two other similar gas turbines with different ambient and operational conditions. The results of the proposed approach implementation demonstrate precise gas turbine sensor fault detection and estimation.

Long Short-Term Memory Network-Based Normal Pattern Group for Fault Detection of Three-Shaft Marine Gas Turbine

Fault detection and diagnosis can improve safety and reliability of gas turbines. Current studies on gas turbine fault detection and diagnosis mainly focus on the case of abundant fault samples. However, fault data are rare or even unavailable for gas turbines, especially newly-run gas turbines. Aiming to realize fault detection with only normal data, this paper proposes the concept of normal pattern group. A group of long-short term memory (LSTM) networks are first used for characterizing the mapping relationships among measurable parameters of healthy three-shaft gas turbines. Experiments show that the proposed method can detect all 13 common gas path faults of three-shaft gas turbines sensitively while remaining low false alarm rate. Comparison experiment with single normal pattern model verifies the necessaries and superiorities of using normal pattern group. Meanwhile, comparison between LSTM network and other methods including support vector regression, single-layer feedforward neural network, extreme learning machine and Elman recurrent neural network verifies the superiorities of LSTM network in fault detection. Furthermore, comparison experiment with four common one-class classifiers further verifies the superiorities of the proposed method. This also indicates the superiorities of data-driven methods and gas turbine principle fusion to some extent.

Fault diagnosis of gas turbine based on matrix capsules with EM routing

The fault detection and diagnosis of a gas turbine is of great significance for guaranteeing the complicated dynamic systems working normally and safely. Most of the existing fault diagnosis methods, based on convolutional neural networks (CNN), have certain limitations in extracting correlations of multi-channel data features. The accuracy of fault diagnosis still needs to be improved. In this paper, an approach of fault diagnosis, based on matrix capsules with EM routing, is presented. First of all, three channels data, which respectively represent acceleration, pressure and pulse, are integrated into one image to feed into the network. Secondly, network models based on the matrix capsules start to be trained by using input dataset which contains fault image and normal image. Finally, the pre-trained capsules model is used to diagnose the state of testing data. Besides, to verify the superiority of the algorithm used in this paper, a comparative experiment is implemented between matrix capsule networks and CNN. The results demonstrate that the testing accuracy is 99.995%.

Application of interval type-2 fuzzy logic systems to gas turbine fault diagnosis

Several approaches have been employed for gas turbine Fault Detection and Identification (FDI), since a reliable FDI system with minimum false alarm rate can effectively reduce maintenance cost and downtime. This paper introduces the application of Interval Type-2 Fuzzy Logic Systems (IT2FLSs) to gas turbine fault diagnosis for the first time. The proposed FDI system is composed of a bank of IT2FLSs, trained for state detection and health assessment of an industrial gas turbine at various operating conditions. For this purpose, train and test data are generated by applying mechanical fault signatures to gas turbine’s mathematical model. Fuzzy Rule Base is then developed by means of Interval Type-2 Fuzzy C-Means (IT2FCM) clustering, and parameters of the IT2FLSs are optimized using a metaheuristic algorithm. Finally, the performance of the IT2FL based FDI system is compared to several classification techniques. It is concluded that as a compromise among the objectives of online applicability, accuracy, reliability against measurement uncertainty, incipient fault diagnosis, robustness against abrupt sensor failure and generalization capacity, the proposed method demonstrates a promising performance.

Fault diagnosis of an industrial gas turbine based on the thermodynamic model coupled with a multi feedforward artificial neural networks

In the study presented in this paper, the deterioration in the performance of an industrial gas turbine during the operation design point was simulated by using the thermodynamic principle and a multi feedforward artificial neural networks (MFANN) system. Initially the thermodynamic model was constructed using the components performance map technique, that entailed calculating the operating point which was compliant with the performance map for each component. The various design operation points were generated by changing the engine component’s efficiency or outer environmental conditions and simulating the engine’s performance for each case. The MFANN model was constructed by using these operation points for the training and testing stage. In this way, the two MFANN models were established. The aim of the first model was to calculate the engine’s performance while the second model was used to detect the deterioration of the components of the engine This paper presents a robust fault diagnosis system for gas turbine degradation detection with the aim of improving energy efficiency.

Fault diagnosis of gas turbine based on partly interpretable convolutional neural networks

The fault diagnosis of gas turbines plays a vital role in engine reliability and availability. The data-driven diagnostic model has been verified useful for identifying and characterizing engine degradation. But Convolution Neural Networks (CNN) is considered to perform poorly in the fault diagnosis for time-series signals. And most of the studies do not involve the interpretability of CNN, leading model hard to be optimized and integrated with physical mechanisms. For the fault diagnosis of the gas turbines, strong coupling often exists between gas path faults and sensor faults, making fault diagnosis difficult when both faults occur simultaneously. A novel method is proposed to improve the performance of typical CNN through optimizing the influence of input measurement parameter sequencing. Extreme Gradient Boosting (XGBoost) is used to make the effects of the sequencing on CNN diagnostic accuracy interpretable. In the simulation experiment, the diagnostic accuracy of CNN after optimization is 95.52%, higher than that of conventional CNN (accuracy rate 91.10%), RNN (accuracy rate 94.21%) and other methods. For the analysis of field data, the new method has shown stronger feature extraction ability and can detect typical gas path faults in advance. The new method performs well in precision, stability, and comprehensibility.

Nonlinear robust fault diagnosis of power plant gas turbine using Monte Carlo-based adaptive threshold approach

This paper addresses the robust fault diagnosis of power plant gas turbine as an uncertain nonlinear system using a new adaptive threshold method. In order to determine the bounds of the adaptive threshold and to identify neural network thresholds modelling, an approach based on Monte Carlo simulation is employed. To evaluate the performance of the proposed fault detection method, a fault sensitivity analysis is provided. In addition, the neural network-based estimators are considered to estimate the magnitude of faults according to the values of residuals. The proposed fault diagnosis system is evaluated during different scenarios. The obtained results indicate the high sensitivity, accuracy, and robustness of the proposed method for fault detection and isolation in the nonlinear uncertain systems, even in dealing with small faults.

A novel gas turbine fault diagnosis method based on transfer learning with CNN

A transfer learning method based on CNN and SVM is investigated for gas turbine fault diagnosis. The excellent classification ability of CNNs is attributed to their ability to learn rich feature representations from a large number of annotated samples. This property, however, currently prevents application of CNNs to problems with fewer samples. This paper shows how feature representations learned with CNN on large-scale annotated gas turbine normal dataset can be efficiently transferred to fault diagnosis task with limited fault data. A feature mapping method to extract the feature representations for fault dataset by reusing the internal layers of CNN trained on the normal dataset is designed, and SVM is used for fault diagnosis. The influence of gas turbine performance parameters arrangement order on proposed method is theoretically analyzed. Finally, the proposed method is validated by the real-life operation data of a gas turbine sample fleet. The experimental results show that despite difference in the two datasets, the transferred feature representations lead to significant improved results for fault diagnosis as well as obviously weaken the individual difference and data noises. The experimental results also confirm that the proposed method has excellent ability for fault diagnosis under small sample condition.