Future failure rate prediction for transformers in power systems

• Main Author: Tan, Jia Le
• Other Authors: Hu Guoqiang
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/158061
• Institution: Nanyang Technological University

Transformers are an integral part of power systems, which means that transformer failure is disruptive to the power grid. Failure prediction is an important tool to estimate when these failures are likely to happen, which informs the replacement of transformers before a failure event. By applying survival analysis methods to transformer data, patterns that are associated with transformer failure can be identified. In particular, the brand of transformer is applied to a Cox model to understand if certain brands have similar hazard functions. The coefficients of these hazard functions are run through a clustering algorithm to calculate the grouping of brands. These groups will serve to simplify the data in future analysis.