基于零序电流波形区间斜率曲线的配电网高阻接地故障检测

配电网高阻接地故障由于故障电流微弱而难以被传统保护装置检测并排除,故障长时间存在会增加发生火灾和人身安全事故的风险。针对现有方法在面对不同接地介质等故障场景时,因故障特征的差异性及噪声干扰而导致的整定阈值难设定和检测可靠性不足等问题,基于对10 k V电网现场实测高阻故障波形的分析,采用最小二乘线性拟合描述波形非线性畸变,提出基于零序电流波形区间斜率曲线的故障检测方法。同时,采用格拉布斯(Grubbs)法抑制不规则波形畸变对区间斜率曲线的干扰,以进一步保证算法对故障特征的正确提取。所提方法具有较强的抗噪声能力,数值仿真和现场实测故障数据验证了算法的可靠性。

Detection of High Impedance Grounding Fault in Distribution Network Based on Interval Slope Curves of Zero-sequence Current

High impedance grounding fault in the distribution network can hardly be detected and cleared by the traditional protective relays due to its weak fault current. The persistent existence of the fault increases the risks of fire hazards and human safety. Aiming at the problems of hard-determined setting threshold and insufficient detection reliability caused by the differences of fault characteristics and noise interference when the existing methods are facing different grounding dielectrics and fault scenarios, a fault detection method is proposed based on interval slope curves of zero sequence current according to the curve analysis of high impedance fault field test in a 10 kV power grid, which uses least squares linear fitting to describe the nonlinear distortion of curves. Meanwhile, Grubbs method is proposed to suppress the interference to interval slopes curves from the irregular waveform distortions and to further guarantee the proper extraction of the fault characteristics. The proposed method possesses strong immunity against noise and the reliability is verified by numerical simulation and field measured fault data.