Research on Transformer Condition-based Maintenance Optimization Based on Live Detection

At present, the maintenance strategy for transformers mainly focus on the optimization of maintenance mode and time, while the influence of state parameter detection mode on power grid is often ignored. Considering the popularization and application of live detection technology, an optimization method of transformer state maintenance strategy based on live detection is proposed, which considers the economy and reliability. Firstly, on the basis of defining the key detection parameters of power outage test, an alternative analysis model based on online oil chromatographic data is constructed for replacing power outage test parameters. And then the average value of absolute error is introduced as an evaluation index to determine whether to replace or extend the power outage test, which is called the formulation of optimal scheme. Finally, a quantitative model of acquisition risk of state parameters and transformer failure risk is established with the comprehensive optimization of reliability and economy as decision variables. The example of maintenance decision-making for a 220kV transformer shows that the state maintenance based on live detection can reduce the maintenance cost on the principle of ensuring safe and reliable operation of transformer.