Study on condition pre-warning method of power transformer based on load time series model

Power transformer is one of key equipment in power system and its normal operation guarantees reliability and safety of power transmission. In order to keep power transformer in good condition, regular inspection and maintenance is needed after equipment put into service. Various preventive test, on-line monitoring and portable test are must, but condition evaluation and fault diagnosis for power transformer still is kind of art, which normally depend on standards, guideline or accumulated experiences on-site. In this paper, insulating condition is evaluated by equipment's operating load, and time series model is established to analyze load characteristics. After outlier value extraction and data reduction, two typical load growth modes have been proposed. Results show that the proposed load time series model could be applied to equipment operating status trend analysis. This approach provides a new idea for power transformer insulation pre-warning by making use of power transformer load characteristic analysis.