Fuzzy sets theory cannot describe the neutrality degreeof data, which has largely limited the objectivity of fuzzy time seriesin uncertain data forecasting. With this regard, a multi-factor highorderintuitionistic fuz...Fuzzy sets theory cannot describe the neutrality degreeof data, which has largely limited the objectivity of fuzzy time seriesin uncertain data forecasting. With this regard, a multi-factor highorderintuitionistic fuzzy time series forecasting model is built. Inthe new model, a fuzzy clustering algorithm is used to get unequalintervals, and a more objective technique for ascertaining membershipand non-membership functions of the intuitionistic fuzzy setis proposed. On these bases, forecast rules based on multidimensionalintuitionistic fuzzy modus ponens inference are established.Finally, contrast experiments on the daily mean temperature ofBeijing are carried out, which show that the novel model has aclear advantage of improving the forecast accuracy.展开更多
时间序列插补旨在根据现有数据填补缺失值以恢复数据的完整性.目前基于RNN的插补方法存在较大的误差,并且增加网络层数容易出现梯度爆炸和消失问题,而基于GAN和VAE的插补方法经常面临训练困难和模式崩溃的挑战.为解决上述问题,本文提出...时间序列插补旨在根据现有数据填补缺失值以恢复数据的完整性.目前基于RNN的插补方法存在较大的误差,并且增加网络层数容易出现梯度爆炸和消失问题,而基于GAN和VAE的插补方法经常面临训练困难和模式崩溃的挑战.为解决上述问题,本文提出了一种基于扩散与时频注意力的时间序列插补模型DTFA (diffusion model and time-frequency attention),通过反向扩散实现从高斯噪声中重建缺失数据.具体而言,本研究利用多尺度卷积模块与二维注意力机制捕获时域数据中的时间依赖性,并利用MLP与二维注意力机制学习频域数据的实部与虚部信息.此外,本研究通过线性插补模块以对现有的观测数据进行初步的数据增强,从而更好地指导模型的插补过程.最后,本研究通过最小化真实噪声与估计噪声的欧氏距离来训练噪声估计网络,并利用反向扩散实现对时序数据的缺失插补.本研究的实验结果表明, DTFA在ETTm1、WindPower和Electricity这3个公开数据集上的插补效果均优于近年主流的基线模型.展开更多
Electric power infrastructure has recently undergone a comprehensive transformation from electromagnetics to semiconductors. Such a development is attributed to the rapid growth of power electronic converter applicati...Electric power infrastructure has recently undergone a comprehensive transformation from electromagnetics to semiconductors. Such a development is attributed to the rapid growth of power electronic converter applications in the load side to realize energy conservation and on the supply side for renewable generations and power transmissions using high voltage direct current transmission. This transformation has altered the fundamental mechanism of power system dynamics, which demands the establishment of a new theory for power system control and protection. This paper presents thoughts on a theoretical framework for the coming semiconducting power systems.展开更多
连接界面上存在的跨尺度、多物理场和非线性行为是引起结构复杂非线性动力学的主要原因.由于连接界面的力学行为的复杂性,以及难以对连接界面进行直接试验观测,连接界面的力学建模一直是非常具有挑战性的科学问题.本文首先从分析结合面...连接界面上存在的跨尺度、多物理场和非线性行为是引起结构复杂非线性动力学的主要原因.由于连接界面的力学行为的复杂性,以及难以对连接界面进行直接试验观测,连接界面的力学建模一直是非常具有挑战性的科学问题.本文首先从分析结合面的跨尺度物理机理入手,将名义的光滑平面视作凹凸不平的粗糙面,考虑单个微凸体的黏滑摩擦行为,建立接触载荷与变形的非线性关系,然后采用GW(Greenwood and Williamson)模型数理统计方法建立整个粗糙界面的跨尺度力学模型,并与公开文献中试验结果进行对比.考虑连接界面具有典型非线性特征,提出一种改进的Iwan唯象模型,利用精细有限元方法获得非线性特征结果,采用系统辨识理论建立连接结构的降阶力学模型,并利用有限元结果进行模型验证.结果表明,本文提出的粗糙界面跨尺度模型在法向载荷较小时与试验结果吻合较好,改进的Iwan模型能够较好描述连接界面的非线性特征,并与有限元结果吻合较好.展开更多
基金supported by the National Natural Science Foundation of China(61309022)
文摘Fuzzy sets theory cannot describe the neutrality degreeof data, which has largely limited the objectivity of fuzzy time seriesin uncertain data forecasting. With this regard, a multi-factor highorderintuitionistic fuzzy time series forecasting model is built. Inthe new model, a fuzzy clustering algorithm is used to get unequalintervals, and a more objective technique for ascertaining membershipand non-membership functions of the intuitionistic fuzzy setis proposed. On these bases, forecast rules based on multidimensionalintuitionistic fuzzy modus ponens inference are established.Finally, contrast experiments on the daily mean temperature ofBeijing are carried out, which show that the novel model has aclear advantage of improving the forecast accuracy.
文摘时间序列插补旨在根据现有数据填补缺失值以恢复数据的完整性.目前基于RNN的插补方法存在较大的误差,并且增加网络层数容易出现梯度爆炸和消失问题,而基于GAN和VAE的插补方法经常面临训练困难和模式崩溃的挑战.为解决上述问题,本文提出了一种基于扩散与时频注意力的时间序列插补模型DTFA (diffusion model and time-frequency attention),通过反向扩散实现从高斯噪声中重建缺失数据.具体而言,本研究利用多尺度卷积模块与二维注意力机制捕获时域数据中的时间依赖性,并利用MLP与二维注意力机制学习频域数据的实部与虚部信息.此外,本研究通过线性插补模块以对现有的观测数据进行初步的数据增强,从而更好地指导模型的插补过程.最后,本研究通过最小化真实噪声与估计噪声的欧氏距离来训练噪声估计网络,并利用反向扩散实现对时序数据的缺失插补.本研究的实验结果表明, DTFA在ETTm1、WindPower和Electricity这3个公开数据集上的插补效果均优于近年主流的基线模型.
基金This work was supported in part by the National Basic Research Program of China (973 Program) (Grant No. 2012CB215100), and the Major Program of the National Natural Science Foundation of China (Grant No. 51190104).
文摘Electric power infrastructure has recently undergone a comprehensive transformation from electromagnetics to semiconductors. Such a development is attributed to the rapid growth of power electronic converter applications in the load side to realize energy conservation and on the supply side for renewable generations and power transmissions using high voltage direct current transmission. This transformation has altered the fundamental mechanism of power system dynamics, which demands the establishment of a new theory for power system control and protection. This paper presents thoughts on a theoretical framework for the coming semiconducting power systems.
文摘连接界面上存在的跨尺度、多物理场和非线性行为是引起结构复杂非线性动力学的主要原因.由于连接界面的力学行为的复杂性,以及难以对连接界面进行直接试验观测,连接界面的力学建模一直是非常具有挑战性的科学问题.本文首先从分析结合面的跨尺度物理机理入手,将名义的光滑平面视作凹凸不平的粗糙面,考虑单个微凸体的黏滑摩擦行为,建立接触载荷与变形的非线性关系,然后采用GW(Greenwood and Williamson)模型数理统计方法建立整个粗糙界面的跨尺度力学模型,并与公开文献中试验结果进行对比.考虑连接界面具有典型非线性特征,提出一种改进的Iwan唯象模型,利用精细有限元方法获得非线性特征结果,采用系统辨识理论建立连接结构的降阶力学模型,并利用有限元结果进行模型验证.结果表明,本文提出的粗糙界面跨尺度模型在法向载荷较小时与试验结果吻合较好,改进的Iwan模型能够较好描述连接界面的非线性特征,并与有限元结果吻合较好.