基于广义基尼指数和脉冲神经网络的航空交流 串联电弧故障检测

航空电缆在振动作用下很容易产生电连接器松动、线束断裂等情况，从而引起交流电弧故障。针对交流串联电弧故 障时频域特征不明显而引起的故障检测问题，提出了一种基于广义基尼指数(generalized Gini indices,GGI)和脉冲神经网络 (spiking neural network,SNN)的电弧故障检测方法。首先，提出用广义基尼指数对试验数据电流波形进行分析；其次判断正 常周期和故障周期下的数值差距，然后与时域特征指标裕度、峭度、脉冲因子相比，所提指数对电流波形周期故障判断更准 确；最后，将广义基尼指数转变成特征值，代入到积分泄漏发放(leaky integrate-and-fire,LIF)模型进行训练，进一步提高方法 的普适性。试验结果表明，该方法能够快速有效地检测航空交流串联电弧故障。

Aviation AC series arc fault detection based on generalized Gini indices and spiking neural network

Aviation cables are prone to loose electrical connectors and broken wiring harnesses under vibration,leading to AC arc faults.A fault detection method for AC series arc faults based on generalized gini exponent and pulse neural network is proposed to address the issue of unclear frequency domain characteristics during fault detection.Firstly,it is proposed to use the generalized Gini index to analyze the current waveform of the test data.Secondly,the numerical difference between normal and fault cycles is determined.Then,compared with the time-domain characteristic indicators such as margin,kurtosis,and pulse factor,the proposed index is more accurate in identifying periodic faults in the current waveform.Finally,the generalized Gini index is transformed into eigenvalues and substituted into the LIF model for training,further improving the universality of the method.The experimental results indicate that this method can quickly and effectively detect aviation AC series arc faults.