摘要
击穿电压是表征发电机定子线棒主绝缘老化状态的特征参量,是影响发电机寿命的直接原因,笔者设计了额定电压为8.5 kV的环氧云母绝缘板材并进行加速热老化试验,通过分析主绝缘板材特征参量的变化趋势,结合击穿电压与特征参量有效性判别,得到能有效表征主绝缘击穿电压的实用特征参量组,并采用BP神经网络滚动式预测方式对主绝缘击穿电压预测,结果表明,该特征参数组Qmax、△C、C、△tanδ、tanδ能有效地用于主绝缘击穿电压预测。
Breakdown voltage is a characteristic parameter for representing main insulation aging of generator stator bar, and is the direct factor affecting generator service life. In this study, a sample of epoxy mica sheet with rated voltage of 8.5 kV was designed to simulate the insulation system of the 8.5 kV generator stator bar.Accelerated thermal aging experiment at 220 ℃ was conducted to obtain the practical characteristic parameters by analyzing the change trend of the insulation sheet＇s characteristic parameters and combining with the validity analysis of breakdown voltage and characteristic parameters. It is shown that the characteristic parameters, Qmax,△C,C, △tanδ, and tanδ can be used to effectively predict breakdown voltage by means of BP neural networks.
出处
《高压电器》
CAS
CSCD
北大核心
2015年第3期105-110,116,共7页
High Voltage Apparatus
关键词
击穿电压
特征参量
主绝缘
breakdown voltage
characteristic parameter
main insulation
