CAJ | 학술논문

将多源信息特征融合应用于光伏并网逆变器故障诊断。选择电网侧电流和重要桥臂电压为融合对象，利用小波变换对数据进行预处理和特征提取，通过神经网络对特征值进行训练，得到故障诊断结果。仿真结果表明，采用多源信息特征融合可有效提高光伏并网逆变器故障诊断的精度。
장다원신식특정융합응용우광복병망역변기고장진단。선택전망측전류화중요교비전압위융합대상，이용소파변환대수거진행예처리화특정제취，통과신경망락대특정치진행훈련，득도고장진단결과。방진결과표명，채용다원신식특정융합가유효제고광복병망역변기고장진단적정도。
The paper applies multi-source feature information fusion to fault diagnosis of photovoltaic grid system in-verter. Fusion objects are chosen as grid side current and important bridge arm voltage, while data preprocessing and feature extraction are performed using wavelet transform. Through the training of feature values with neural network, fault diagnosis results are obtained. Simulation results show that the multi-source information fusion can effectively im-prove the accuracy of photovoltaic grid inverter.