This paper presents a novel, automated, generative adversarial networks (GAN) based synthetic feeder generation mechanism, abbreviated as FeederGAN. FeederGAN digests real feeder models represented by directed graphs via a deep learning framework powered by GAN and graph convolutional networks (GCN). From power system feeder model input files, device connectivity is mapped to the adjacency matrix while device characteristics such as circuit types (i.e., 3-phase, 2-phase, and 1-phase) and component attributes (e.g., length and current ratings) are mapped to the attribute matrix. Then, Wasserstein distance is used to optimize the GAN and GCN is used to discriminate the generated graph from the actual. A greedy method based on graph theory is developed to reconstruct the feeder from the generated adjacency and attribute matrix. Our results show that the generated feeders resemble the actual feeder in both topology and attributes verified by visual inspection and by empirical statistics obtained from actual distribution feeders.
翻译:本文介绍了一种新型、自动化、基因化的对抗性合成种子网络(GAN)生成机制,缩写为FeelerGAN。 FeterGAN摘要通过由GAN和图图变网络驱动的深层学习框架,通过GCN和图变网络(GCN),以定向图形为代表的真正种子模型模型模型。从动力系统种子模型输入文档中,设备连接映射到相近矩阵,而电路类型(即3级、2级和1级)和组成部分属性(例如长度和当前评级)等设备特性被映射到属性矩阵中。然后,用瓦塞尔斯坦距离优化GAN和GCN用于区分实际生成的图形。一种基于图形理论的贪婪方法,从生成的相近性和属性矩阵中重建饲料。我们的结果表明,生成的种子器与通过直观检查和从实际分发源获得的经验性统计数据所证实的表层和属性中的实际种子。