基于溶解气体分析的电力变压器在线监测与诊断技术

On-line monitoring and diagnosis of power transformer based on dissolved gas analysis

油中溶解气体分析是评估变压器绝缘状态和分析变压器绝缘故障的重要依据。在介绍基于油中溶解气体分析的变压器在线监测与诊断技术原理的基础上,进一步总结了当前在线监测与诊断系统的研究现状及存在的问题:系统主要是在线监测油中气体成分及超阈值报警,而对故障性质、种类、定位及发展趋势预测等诊断尚不具备或很不完备,故障诊断需靠离线分析;因缺乏相应的理论指导,在线监测数据在正常情况下应满足的规律还很模糊,装置自身的精确度不能保证。对今后的研究提出了展望,指出如何通过对在线数据的挖掘找到变压器故障诊断的相应理论依据是进一步研究的核心。

The DGA（Dissolved Gas Analysis） is essential for the insulation state assessment and the fault diagnosis of transformers. The principle of the DGA-based on-line monitoring and diagnosis system for transformers is first introduced in which its present research status and problems are summarized. The on-line system mainly monitors gases in oil and alarms when exceeding. It acquires poor or even no information about the fault character,the type, the location and the development trend,the system depends on the off- line analysis for diagnosis. Rules for on-line data in the normal operation is still fuzzy so far due to lacking of associated theories, and the precision of the equipment itself can not be ensured. Prospects are finally presented,in which how to find out a fundamental theory for the transformer diagnosis through the on-line data mining is the core.