LTD多间隙气体开关低概率自放电估计

Estimation of low probability pre-fire of multi-gap gas switch for linear transformer driver

  • 摘要: 开关触发前发生的低概率自击穿故障,是制约直线变压器驱动源(LTD)技术发展和工程应用的主要因素。针对多间隙气体开关的自放电概率展开研究,设计了简化结构的气体开关来测量更高气压下的自击穿电压;推导了考虑间隙分压不均的开关自击穿电压累积分布函数,计算LTD工作气压下的多间隙开关自击穿电压分布特性并进行实验验证;通过超越阈值方法对低工作系数下的小概率自放电事件进行威布尔分布拟合,得到自放电概率估计值;对气体开关开展了万发次考核实验,得到自放电概率测量值。结果表明:开关的自击穿电压分散性随着气压升高而增加,测量低气压下的自击穿电压无法外推得到实际工作高气压条件下的自放电概率;开关工作系数为65%和60%时,气体开关的自放电概率估计值为1.44 \times 10^ - 43.01 \times 10^ - 5,与开关万发次考核的实验结果相近,验证了估计方法的可行性。

     

    Abstract: The pre-fire of gas switches is the unexpected breakdown at triggered mode, which is the main factor restricting the development and engineering application of linear transformer driver (LTD). In this paper, the dispersion of breakdown voltage as well as pre-fire rate of multi-gap gas switches under typical working pressures of LTD was investigated and reported. A simplified gas switch was designed to measure self-breakdown voltage at higher gas pressures, and the cumulative distribution function of self-breakdown voltage considering uneven gap voltage distribution was derived. The self-breakdown voltage distribution characteristics of multi-gap switches under LTD working pressure were calculated and experimentally verified. The estimated pre-fire rate was obtained by using peak-over-threshold method to fit the Weibull distribution of small probability pre-fire events under low working coefficients. The measured pre-fire rate was obtained by a 10 000 triggering test. The results show that the dispersion of the self-breakdown voltage of the switch increased with air pressure. Measuring the self-breakdown voltage at low air pressure can not extrapolate the pre-fire rate at high pressure under actual working conditions. When the working coefficients of the gas switch were 65% and 60%, the estimated pre-fire rates were 1.44 \times 10^ - 4 and 3.01 \times 10^ - 5, which were close to the experimental results of the 10 000 triggering test, verifying the feasibility of the estimation method.

     

/

返回文章
返回