A Lecturer from the Department of Mathematics Publishes a Global Research Paper.
Asst. Lect. Ali Muhammad Hussein from the Department of Mathematics published a scientific paper titled: A New Transformer Fault Diagnosis Method Based on an Improved Fuzzy C-means Clustering Algorithm
Publisher: IEEE Journal
The aim of the research was to explain the proper performance of transformers and the proper operation of the power sector. Accurate diagnosis of problems associated with transformers is essential for their maintenance and continued efficient operation. Several methods are used for this purpose, including the three-ratio method. This method is commonly used to diagnose faults by analyzing the gas emissions of oil-immersed transformers.
The research included the three-ratio method, which uses the Fuzzy C-Means clustering (FCM) algorithm, a suitable approach to address this problem in the traditional three-ratio method. However, the FCM algorithm suffers from a random initial selection problem. This paper addresses the shortcomings of the random initial selection of the FCM algorithm to improve its performance in accurately diagnosing transformer faults. For evaluation, we compared the proposed method with the traditional FCM model and the improved FCM model (I-FCM).
The study conducted several experiments on some transformer fault diagnosis sample sets using the popular IRIS dataset. The results showed that the proposed method generally outperformed the traditional I-FCM and FCM models.
