基于混沌多项式展开的地磁感应电流不确定性研究

Study on Uncertainty of Geomagnetically Induced Currents Based on Polynomial Chaos Expansion

地磁感应电流(GIC)会引起变压器直流偏磁,其次生衍生效应可能威胁电力设备和电网的安全。研究磁暴期间大地电导率参数变化引起的GIC不确定性对GIC的评估和防御具有重要意义。基于混沌多项式展开(PCE)方法,以一维水平分层大地电导率为输入变量,在原有电场及GIC计算的基础上构造了二者的混沌多项式展开式。针对三层大地电导率模型和Benchmark算例,利用所构造的混沌多项式对磁暴期间感应地电场和电网中各支路GIC进行了不确定度量化分析,得到了GIC最大值的95%置信区间、均值和方差等统计量。通过与蒙特卡洛法(MC)的计算结果对比,验证了PCE方法的有效性,且PCE方法计算效率远高于MC方法。

Geomagnetic induction current(GIC)can cause transformer DC bias,whose secondary derivative effect may threaten the safety of power equipment and power grid. Studying the GIC uncertainty caused by the earth conductivity variation during magnetic storms is of great significance to the assessment and prevention of GIC. Based on the polynomial chaos expansion (PCE)method,assuming the conductivity of one dimensional earth model as input variables,the chaos polynomial expansion of induced electric field and GIC are constructed on the basis of the original electric field and the GIC calculation. For the threelayer earth conductivity model and the Benchmark example,the uncertainty analysis for induced electric field and the GIC of each branch of the power grid are carried out by using the derived polynomial chaos expansions and the 95% confidence interval,mean and variance statistics of the maximum GIC are obtained. Comparison between the PCE method and the Monte Carlo(MC)method shows the validity and the efficiency of the PCE method.