半监督学习可借助有标签和部分无标签样本数据来构建电网暂态稳定评估模型,有效利用输入样本数据,可提高电网暂态稳定评估准确率,为此提出基于半监督近似流形支持向量机(manifold proximal support vector machine,MPSVM)的暂态稳定评...半监督学习可借助有标签和部分无标签样本数据来构建电网暂态稳定评估模型,有效利用输入样本数据,可提高电网暂态稳定评估准确率,为此提出基于半监督近似流形支持向量机(manifold proximal support vector machine,MPSVM)的暂态稳定评估方法。首先,在MPSVM的正则项中引入判别变量,可最大限度捕捉样本数据内部的几何信息,并通过最大距离理论表征电力系统稳定类和不稳定类之间的差异,进而转化为求解特征值问题;然后,采用贝叶斯非线性分层模型确定最优参数,可进一步提高评估准确率;最后,采用IEEE 39标准系统和鞍山电网的仿真分析验证所提评估模型的有效性和准确性。展开更多
Sixty indexes, ranging from social development and economy increasing to ecological protection and transportation system construction, are chosen to construct an index system of adaptability evaluation on transportati...Sixty indexes, ranging from social development and economy increasing to ecological protection and transportation system construction, are chosen to construct an index system of adaptability evaluation on transportation system and urban development. A synthesis model with nonlinear characteristic is proposed, and the corresponding procedure is presented based on an improved AHP model, which utilizes the SVD method to improve the required precision of matrix with acceptable consistency weight and detect and amend the significant elements via distance and proximity so as to improve the whole consistency. Taking Shandong Province for the case study, experimental results indicate that the adaptability degree of the system is in an increasing state, which provides robust and effective support for decision-makers working on a range of problems and in various circumstances.展开更多
基金Sponsored by the Key Projects of the National Science & Technology Pillar Program in the Eleventh Five-year Plan Period (Grant No.2006BAJ18B01)
文摘Sixty indexes, ranging from social development and economy increasing to ecological protection and transportation system construction, are chosen to construct an index system of adaptability evaluation on transportation system and urban development. A synthesis model with nonlinear characteristic is proposed, and the corresponding procedure is presented based on an improved AHP model, which utilizes the SVD method to improve the required precision of matrix with acceptable consistency weight and detect and amend the significant elements via distance and proximity so as to improve the whole consistency. Taking Shandong Province for the case study, experimental results indicate that the adaptability degree of the system is in an increasing state, which provides robust and effective support for decision-makers working on a range of problems and in various circumstances.
