Start-up vibration analysis for novelty detection on industrial gas turbines

Zhang, Yu; Cruz-Manzo, Samuel; Latimer, Anthony

doi:10.1109/indel.2016.7797800

Cited by 6 publications

(5 citation statements)

References 14 publications

(13 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Datasets from the identified transient operation can also be used for fault detection through start-up analysis and shutdown analysis and during load changes [6,24,25], which is not included in the current paper. The most relevant features are then extracted from the steady-state data and a statistical "fingerprint" for the extracted features is obtained through the application of GMMOC.…”

Section: Methodsmentioning

confidence: 99%

“…VBGMMs are able to classify steady-state operation that can occur under full-or part-load conditions (e.g., 50% load). Additionally therefore, as well as identifying steadystate operation, the remaining data, including that associated with start-ups, shutdowns, and load changing conditions, is also naturally separated and it can therefore be used as a preanalysis tool for alternative dynamic scenarios, as reported in [24,25], for instance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Detection of Emerging Faults on Industrial Gas Turbines Using Extended Gaussian Mixture Models

Zhang

Bingham

Martínez-García

et al. 2017

International Journal of Rotating Machinery

Self Cite

View full text Add to dashboard Cite

This paper extends traditional Gaussian mixture model (GMM) techniques to provide recognition of operational states and detection of emerging faults for industrial systems. A variational Bayesian method allows a GMM to cluster with its mixture components to facilitate the extraction of steady-state operational behaviour; this is recognised as being a primary factor in reducing the susceptibility of alternative prognostic/diagnostic techniques, which would initiate false-alarms resulting from control set-point and load changes. Furthermore, a GMM with an outlier component is discussed and applied for direct novelty/fault detection. An advantage of the variational Bayesian method over traditional predefined thresholds is the extraction of steady-state data during both full- and part-load cases, and a primary advantage of the GMM with an outlier component is its applicability for novelty detection when there is a lack of prior knowledge of fault patterns. Results obtained from the real-time measurements on the operational industrial gas turbines have shown that the proposed technique provides integrated preprocessing, benchmarking, and novelty/fault detection methodology.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Detection of Emerging Faults on Industrial Gas Turbines Using Extended Gaussian Mixture Models

Zhang

Bingham

Martínez-García

et al. 2017

International Journal of Rotating Machinery

Self Cite

View full text Add to dashboard Cite

show abstract

“…There are methods based on creating a graphical representation of the physical system, such as the bond graph [17]. This graphical representation increases one's insight into systems behaviour and understanding of the energy transfer between the inner components.…”

Section: Related Workmentioning

confidence: 99%

Monitoring the Gas Turbine Start-Up Phase on a Platform Using a Hierarchical Model Based on Multi-Layer Perceptron Networks

Niksa-Rynkiewicz

Witkowska

Głuch

et al. 2022

Polish Maritime Research

View full text Add to dashboard Cite

Very often, the operation of diagnostic systems is related to the evaluation of process functionality, where the diagnostics is carried out using reference models prepared on the basis of the process description in the nominal state. The main goal of the work is to develop a hierarchical gas turbine reference model for the estimation of start-up parameters based on multi-layer perceptron neural networks. A functional decomposition of the gas turbine start-up process was proposed, enabling a modular analysis of selected parameters of the process. Real data sets obtained from observations of the turbo-generator set located on a North Sea platform were used.

show abstract

“…They finally tested ALSTOM GT13E2 gas turbine to validate their proposed model. Regarding novelty detection in industrial gas turbines, Zhang et al [10] focused on analyzing start-up vibration. To this end, a vibration signature obtained from accurate measurements based on a neuro-fuzzy system was necessary.…”

Section: Introductionmentioning

confidence: 99%

Modeling and Control of the Starter Motor and Start-Up Phase for Gas Turbines

2019

View full text Add to dashboard Cite

Improving the performance of industrial gas turbines has always been at the focus of attention of researchers and manufacturers. Nowadays, the operating environment of gas turbines has been transformed significantly respect to the very fast growth of renewable electricity generation where gas turbines should provide a safe, reliable, fast, and flexible transient operation to support their renewable partners. So, having a reliable tools to predict the transient behavior of the gas turbine is becoming more and more important. Regarding the response time and flexibility, improving the turbine performance during the start-up phase is an important issue that should be taken into account by the turbine manufacturers. To analyze the turbine performance during the start-up phase and to implement novel ideas so as to improve its performance, modeling, and simulation of an industrial gas turbine during cold start-up phase is investigated this article using an integrated modular approach. During this phase, a complex mechatronic system comprised of an asynchronous AC motor (electric starter), static frequency converter drive, and gas turbine exists. The start-up phase happens in this manner: first, the clutch transfers the torque generated by the electric starter to the gas turbine so that the turbine reaches a specific speed (cranking stage). Next, the turbine spends some time at this speed (purging stage), after which the turbine speed decreases, sparking stage begins, and the turbine enters the warm start-up phase. It is, however, possible that the start-up process fails at an intermediate stage. Such unsuccessful start-ups can be caused by turbine vibrations, the increase in the gradients of exhaust gases, or issues with fuel spray nozzles. If, for any reason, the turbine cannot reach the self-sustained speed and the speed falls below a certain threshold, the clutch engages once again with the turbine shaft and the start-up process is repeated. Consequently, when modeling the start-up phase, we face discontinuities in performance and a system with variable structure owing to the existence of clutch. Modeling the start-up phase, which happens to exist in many different fields including electric and mechanical application, brings about problems in numerical solutions (such as algebraic loop). Accordingly, this study attempts to benefit from the bond graph approach (as a powerful physical modeling approach) to model such a mechatronic system. The results confirm the effectiveness of the proposed approach in detailed performance prediction of the gas turbine in start-up phase.

show abstract

Start-up vibration analysis for novelty detection on industrial gas turbines

Cited by 6 publications

References 14 publications

Detection of Emerging Faults on Industrial Gas Turbines Using Extended Gaussian Mixture Models

Detection of Emerging Faults on Industrial Gas Turbines Using Extended Gaussian Mixture Models

Monitoring the Gas Turbine Start-Up Phase on a Platform Using a Hierarchical Model Based on Multi-Layer Perceptron Networks

Modeling and Control of the Starter Motor and Start-Up Phase for Gas Turbines

Contact Info

Product

Resources

About