A Fault Hierarchical Diagnosis Method for the Generator Excitation System Based on Health Assessment Feedback

In order to provide a comprehensive solution for early and accurate fault diagnosis in generator excitation systems, this paper presents a fault hierarchical discrimination method based on health degree assessment feedback. Firstly, a total of 6 layers is built to constitute the hierarchical diagnosis structure, which are threshold decision layer, status warning layer, fault tree diagnosis layer, expert advice layer, health assessment layer and evaluation feedback layer, respectively. Through monitoring the condition parameters of generator excitation system in real time, the threshold criterion layer and the state warning layer are used for prompting the occurrence information and the warning information of Class II fault. Then, on the basis of fault tree diagnosis layer, the prompted fault is refined into Class III or IV fault, while the corresponding diagnostic opinions and countermeasures are also obtained according to the expert advice layer. Moreover, the model adjustment principle of evaluation feedback layer is investigated as well as the model adjustment strategies of threshold decision layer and status warning layer so that the method performances can adapt to the health assessment result. Finally, the effectiveness and feasibility of the proposed method have been validated through experiments on the RTDS testbench.