Abstract

Gas turbines have played a key role in aeronautical industry, power generation and as mechanical drives for pumps and compressors. To monitor the efficiency and reliability of gas turbines, the employment and improvement of a remote condition monitoring and fault diagnostic system is of great importance. This paper introduces a scheme of condition monitoring for gas turbines, including sensor validation, steady-state and transient operation discrimination, and novelty or fault detection. Moreover, the methodologies are categorised as knowledge-based rules, signal processing-based techniques and model-based approaches. Among them, model-based approaches further take account of white-box, black-box and grey-box modelling techniques. The strengths and weaknesses of the methods are discussed, and future work in this research area is also recommended.