Transformer aging failure rate evaluation method based on evidence theory for operational risk assessment

Failure rate is a basic parameter in the outage model of the component in power systems. A novel aging failure rate evaluation method based on evidence theory is proposed. Transformer is taken as an example to illustrate the proposed method, because the transformers have been widely used online condition monitors. The operational states of a transformer can be classified into four degrees according to IEEE standards. An evaluation method is proposed to identify the operational state of a transformer by combining the data collected from condition monitors, which is based on evidence theory. Hidden Markov Models are used to model the aging process of a transformer, while the transition rate in the models can be obtained from the historical data. The time-varying aging failure rate function can be derived by solving the Markov state equation. A real example is presented to demonstrate the proposed method reasonable.