Fuzzy Diagnostic Strategy Implementation for Gas Turbine Vibrations Faults Detection: Towards a Characterization of Symptom–fault Correlations

Currently, the challenges of sustainable development in the operational safety of industrial systems continue to grow and increase, given the increasing complexity of these industrial installations. Through the recurrent development of better competitive performance in terms of quality, productivity, flexibility, robustness, and environmental impacts, which is essential in order to offer even more suitable and more accurate solutions to monitoring issues. In order to avoid the dangers that may arise because of the risks and dysfunctional faults of these systems, these risks require mastery of the monitoring systems inherent in this equipment. This work proposes an original approach to gas turbine monitoring, based on the concept of fuzzy logic, with practical implementations of this diagnosis strategy for vibration faults of this rotating machine. The proposed approach makes it possible to monitor the efficiency of the studied gas turbine in real-time and for substantial and substantive improvement in the operation of this turbine and to achieve the objectives in terms of operational safety and in terms of solutions' costs implemented for this type of machine. To this end, first fuzzy models associated with real behavior models of the examined turbine variables are developed, to identify and model the gas turbine dynamics used in a gas transport facility at Hassi R'Mel in southern Algeria, based on a fuzzy approach using experimental operating data from the rotating machine, in order to approximate the variables of this nonlinear system in the form of fuzzy rules and variables. Then, the fault diagnosis strategy of this turbine based on the generation of fault residues is carried out and tested in the presence and absence of disturbances, to validate the effectiveness of this fuzzy approach in terms of sensitivity and robustness of detection, location, and identification of faults. This work shows the efficiency and robustness of the fuzzy gas turbine faults detection strategy and improves the operation of this turbine. Likewise, has increased turbine work time, minimized downtime, and reduced the frequency of unscheduled downtime of this rotating machine.