摘要
针对神经网络应用于电力系统暂态稳定评估存在的误分类问题,将粗糙集理论和神经网络相结合,运用特征矩阵进行属性约简的基础上,应用装袋策略构造集成神经网络分类器来提高分类准确率.在新英格兰10机39节点系统中的应用验证了该分类器的分类准确率较普通神经网络分类器有较大的提高.
Considering that misclassifications often occur while applying ensemble artificial neural networks in transient stability assessment （TSA） of a power system, the paper, based on the attributes reduction by characteristic matrix in rough sets, presents a new ensemble classifier which can improve the reliability of TSA. The classifier is built with the bagging method. Its applications to the New England 10-Machine 39-Bus Power Systems demonstrate its validity for TSA problems of power systems.
出处
《南京工程学院学报（自然科学版）》
2006年第1期7-13,共7页
Journal of Nanjing Institute of Technology(Natural Science Edition)
关键词
人工神经网络
暂态稳定
特征矩阵
装袋策略
集成分类器
artificial neural networks
transient stability assessments （TSA）
characteristic matrix
bagging
ensemble classifier
