The way of neutral point to earth via full compensation arc suppression coil can solve the problem of residual current compensation in coal mine power network effectively. Based on the analysis on the grounding curren...The way of neutral point to earth via full compensation arc suppression coil can solve the problem of residual current compensation in coal mine power network effectively. Based on the analysis on the grounding current detection results of Xieqiao coal mine, the conclusion that harmonic component of grounding current is dominated by higher harmonics with complex harmonic sources in coal mine power network system was obtained. The influences of harmonic source type and fault point position on harmonic voltage and harmonic current were analyzed theoretically. The influences of earthed fault feeder detection result and the estimation errors of parameters to earth on residual current compensation were analyzed. A new thought of residual current prediction and the selections of model method and control method were proposed on this basis. The simulation results prove that harmonic amplitudes of zero sequence voltage and zero sequence current are determined by harmonic source type as well as fault point position in coal mine power network, and also prove that zero sequence voltage detection can avoid the unstable problem of coal mine power network system caused by undercompensation of capacitive current. Finally, the experimental device of full compensation arc suppression coil is introduced.展开更多
基于风电场复杂运行环境和多分支混合集电线路的单相接地故障定位需求,提出一种基于卷积神经网络(convolutional neural network,CNN)和长短期记忆网络(long short term memory networks,LSTM)混合模型(CNN-LSTM)的单相接地故障定位策...基于风电场复杂运行环境和多分支混合集电线路的单相接地故障定位需求,提出一种基于卷积神经网络(convolutional neural network,CNN)和长短期记忆网络(long short term memory networks,LSTM)混合模型(CNN-LSTM)的单相接地故障定位策略。采集故障时零序电流,构建风电场单相接地故障数据集,将CNNLSTM混合模型改进为适合故障测距的预测模型,将该模型应用于在线故障定位。研究结果表明:与CNN和反向传播神经网络(backpropagation neural network,BP)相比,CNN-LSTM混合模型方法定位准确率更高,在不同故障距离和故障电阻情况下均可使用。研究结论为风电场集电线路接地故障定位提供参考。展开更多
微机继电保护装置在变电站复杂的电磁干扰环境下,内部的电子元器件将受到干扰使其不能正常运行,甚至影响到保护的正确动作,对电网的安全稳定运行构成威胁.结合白河变500 k VⅠ母BP-2B母差保护定期检查工作中发现的缺陷,采用对称分量法...微机继电保护装置在变电站复杂的电磁干扰环境下,内部的电子元器件将受到干扰使其不能正常运行,甚至影响到保护的正确动作,对电网的安全稳定运行构成威胁.结合白河变500 k VⅠ母BP-2B母差保护定期检查工作中发现的缺陷,采用对称分量法对采样结果进行分析,确认故障原因是无法采集到零序电流,对可能的故障点进行排查,最终定位故障插件并提出了诊断方法和预防措施.展开更多
基金The financial support from the National Natural Science Foundation of China (No. 51107143)
文摘The way of neutral point to earth via full compensation arc suppression coil can solve the problem of residual current compensation in coal mine power network effectively. Based on the analysis on the grounding current detection results of Xieqiao coal mine, the conclusion that harmonic component of grounding current is dominated by higher harmonics with complex harmonic sources in coal mine power network system was obtained. The influences of harmonic source type and fault point position on harmonic voltage and harmonic current were analyzed theoretically. The influences of earthed fault feeder detection result and the estimation errors of parameters to earth on residual current compensation were analyzed. A new thought of residual current prediction and the selections of model method and control method were proposed on this basis. The simulation results prove that harmonic amplitudes of zero sequence voltage and zero sequence current are determined by harmonic source type as well as fault point position in coal mine power network, and also prove that zero sequence voltage detection can avoid the unstable problem of coal mine power network system caused by undercompensation of capacitive current. Finally, the experimental device of full compensation arc suppression coil is introduced.
文摘基于风电场复杂运行环境和多分支混合集电线路的单相接地故障定位需求,提出一种基于卷积神经网络(convolutional neural network,CNN)和长短期记忆网络(long short term memory networks,LSTM)混合模型(CNN-LSTM)的单相接地故障定位策略。采集故障时零序电流,构建风电场单相接地故障数据集,将CNNLSTM混合模型改进为适合故障测距的预测模型,将该模型应用于在线故障定位。研究结果表明:与CNN和反向传播神经网络(backpropagation neural network,BP)相比,CNN-LSTM混合模型方法定位准确率更高,在不同故障距离和故障电阻情况下均可使用。研究结论为风电场集电线路接地故障定位提供参考。
文摘微机继电保护装置在变电站复杂的电磁干扰环境下,内部的电子元器件将受到干扰使其不能正常运行,甚至影响到保护的正确动作,对电网的安全稳定运行构成威胁.结合白河变500 k VⅠ母BP-2B母差保护定期检查工作中发现的缺陷,采用对称分量法对采样结果进行分析,确认故障原因是无法采集到零序电流,对可能的故障点进行排查,最终定位故障插件并提出了诊断方法和预防措施.
