微型电脑应用

为了高效准确地实现配网故障分类，提出了一种基于卷积神经网络的配网故障分类方法。这种方法可直接对三相电压和电流信号的原始数据样本进行处理，输出包括无故障在内的所有类型的配网故障类型。应用10倍交叉验证方法的测试结果表明，所提方法的准确度为99.52%。在与基于传统机器学习模型的故障分类方法进行了对比实验中，所提方法在准确性和计算效率方面均表现出更好的分类性能。将所提出方法应用在未集成和集成分布式电源的配电系统中，分别达到了99.92%和99.97%的分类精度。

In order to classify efficiently and accurately distribution network fault,a distribution network fault classification method based on convolutional neural network is proposed.The proposed method can directly process the raw data samples of the three-phase voltage and current signals,and output all types of distribution network fault types including no faults.The test results of applying the 10-fold cross-validation method show that the accuracy of the proposed method is 99.52%.In comparison experiments with fault classification methods based on traditional machine learning models,the proposed method shows better classification performance in terms of both accuracy and computational efficiency.The proposed method is applied to power distribution systems with unintegrated and integrated distributed power generation,and achieves classification accuracies of 99.92% and 99.97%.