Software system can be classified into many function modules from the perspective of user. Unified modeling language( UML) class diagram of each function module was extracted,and design characteristic metrics which in...Software system can be classified into many function modules from the perspective of user. Unified modeling language( UML) class diagram of each function module was extracted,and design characteristic metrics which influenced software maintainability were selected based on UML class diagram.Choosing metrics of UML class diagram as predictors,and mean maintenance time of function module was regarded as software maintainability parameter. Software maintainability models were built by using back propagation( BP) neural network and radial basis function( RBF) neural network, respectively and were simulated by MATLAB. In order to evaluate the performance of models,the training results were analyzed and compared with leaveone-out cross-validation and model performance evaluation criterion. The result indicated that RBF arithmetic was superior to BP arithmetic in predicting software maintainability.展开更多
利用疏散星团NGC 188所在天区的1046颗恒星样本的高精度3维(3D)运动学数据(自行和视向速度)测试了DBSCAN(Density-Based Spatial Clustering of Applications with Noise)聚类算法的成员判定效果.为了避免自行和视向速度的单位不一致带...利用疏散星团NGC 188所在天区的1046颗恒星样本的高精度3维(3D)运动学数据(自行和视向速度)测试了DBSCAN(Density-Based Spatial Clustering of Applications with Noise)聚类算法的成员判定效果.为了避免自行和视向速度的单位不一致带来的影响,在数据预处理阶段将3个分量的数据统一标准化至[0,1]区间.利用第k个最近邻点距离方法分析了1046颗恒星样本在标准化无量纲3D速度空间的分布特征,再根据第k个最近邻点距离随k值的变化趋势确定了DBSCAN聚类算法的输入参数(Eps,MinPts),最后利用DBSCAN聚类算法分离出497颗3D运动学成员星.分析结果表明得到的3D运动学成员星是可靠的.展开更多
研究基于CIM_SVG的电网数据模型基础上的配电线路单线图自动布局算法,给出了一种辐射状单线图自动绘图的完整方案。提出一种馈线分级分支的干线-支路模型,结合CIM_SVG数据模型垂直、水平逐级交叉布置馈线各支线;应用面向对象思想建立一...研究基于CIM_SVG的电网数据模型基础上的配电线路单线图自动布局算法,给出了一种辐射状单线图自动绘图的完整方案。提出一种馈线分级分支的干线-支路模型,结合CIM_SVG数据模型垂直、水平逐级交叉布置馈线各支线;应用面向对象思想建立一种支线图元区域模型,通过模型数据判别图形中的极不对称支线以及重叠或交叉区域,进而根据判别的结果对图形的整体对称化处理以及重叠或交叉区域的消除,最终完成整条馈线的绘制。最后以10 k V辐射状配电馈线为例对算法进行验证,结果表明算法能在数秒内完成典型配电网络布局,适用于更广泛的配电网可视化分析应用。展开更多
High-precision proper motions and radial velocities of 1046 stars are used to determine member stars using three-dimensional （3D） kinematics for open clus- ter NGC 188 based on the density-based spatial clustering o...High-precision proper motions and radial velocities of 1046 stars are used to determine member stars using three-dimensional （3D） kinematics for open clus- ter NGC 188 based on the density-based spatial clustering of applications with noise （DBSCAN） clustering algorithm. By implementing this algorithm, 472 member stars in the cluster are obtained with 3D kinematics. The color-magnitude diagram （CMD） of the 472 member stars using 3D kinematics shows a well-defined main sequence and a red giant branch, which indicate that the DBSCAN clustering algorithm is very effective for membership determination. The DBSCAN clustering algorithm can ef- fectively select probable member stars in 3D kinematic space without any assumption about the distribution of the cluster or field stars. Analysis results show that the CMD of member stars is significantly clearer than the one based on 2D kinematics, which al- lows us to better constrain the cluster members and estimate their physical parameters. Using the 472 member stars, the average absolute proper motion and radial velocity are determined to be （PMα, PMδ） = （-2.58 ± 0.22, ＋0.17 ± 0.18） mas yr-1 and Vr = -42.35 ± 0.05 km s-1, respectively. Our values are in good agreement with values derived by other authors.展开更多
Rotor blades in a radial turbine with nozzle guide vanes typically experience harmonic aerodynamic excitations due to the rotor stator interaction. Dynamic stresses induced by the harmonic excitations can result in hi...Rotor blades in a radial turbine with nozzle guide vanes typically experience harmonic aerodynamic excitations due to the rotor stator interaction. Dynamic stresses induced by the harmonic excitations can result in high cycle fatigue(HCF) of the blades. A reliable prediction method for forced response issue is essential to avoid the HCF problem. In this work, the forced response mechanisms were investigated based on a fluid structure interaction(FSI) method. Aerodynamic excitations were obtained by three-dimensional unsteady computational fluid dynamics(CFD) simulation with phase shifted periodic boundary conditions. The first two harmonic pressures were determined as the primary components of the excitation and applied to finite element(FE) model to conduct the computational structural dynamics(CSD) simulation. The computed results from the harmonic forced response analysis show good agreement with the predictions of Singh's advanced frequency evaluation(SAFE) diagram. Moreover, the mode superposition method used in FE simulation offers an efficient way to provide quantitative assessments of mode response levels and resonant strength.展开更多
文摘Software system can be classified into many function modules from the perspective of user. Unified modeling language( UML) class diagram of each function module was extracted,and design characteristic metrics which influenced software maintainability were selected based on UML class diagram.Choosing metrics of UML class diagram as predictors,and mean maintenance time of function module was regarded as software maintainability parameter. Software maintainability models were built by using back propagation( BP) neural network and radial basis function( RBF) neural network, respectively and were simulated by MATLAB. In order to evaluate the performance of models,the training results were analyzed and compared with leaveone-out cross-validation and model performance evaluation criterion. The result indicated that RBF arithmetic was superior to BP arithmetic in predicting software maintainability.
文摘利用疏散星团NGC 188所在天区的1046颗恒星样本的高精度3维(3D)运动学数据(自行和视向速度)测试了DBSCAN(Density-Based Spatial Clustering of Applications with Noise)聚类算法的成员判定效果.为了避免自行和视向速度的单位不一致带来的影响,在数据预处理阶段将3个分量的数据统一标准化至[0,1]区间.利用第k个最近邻点距离方法分析了1046颗恒星样本在标准化无量纲3D速度空间的分布特征,再根据第k个最近邻点距离随k值的变化趋势确定了DBSCAN聚类算法的输入参数(Eps,MinPts),最后利用DBSCAN聚类算法分离出497颗3D运动学成员星.分析结果表明得到的3D运动学成员星是可靠的.
文摘研究基于CIM_SVG的电网数据模型基础上的配电线路单线图自动布局算法,给出了一种辐射状单线图自动绘图的完整方案。提出一种馈线分级分支的干线-支路模型,结合CIM_SVG数据模型垂直、水平逐级交叉布置馈线各支线;应用面向对象思想建立一种支线图元区域模型,通过模型数据判别图形中的极不对称支线以及重叠或交叉区域,进而根据判别的结果对图形的整体对称化处理以及重叠或交叉区域的消除,最终完成整条馈线的绘制。最后以10 k V辐射状配电馈线为例对算法进行验证,结果表明算法能在数秒内完成典型配电网络布局,适用于更广泛的配电网可视化分析应用。
基金supported by the School Foundation of Changzhou University(Grant No.ZMF1002121)
文摘High-precision proper motions and radial velocities of 1046 stars are used to determine member stars using three-dimensional （3D） kinematics for open clus- ter NGC 188 based on the density-based spatial clustering of applications with noise （DBSCAN） clustering algorithm. By implementing this algorithm, 472 member stars in the cluster are obtained with 3D kinematics. The color-magnitude diagram （CMD） of the 472 member stars using 3D kinematics shows a well-defined main sequence and a red giant branch, which indicate that the DBSCAN clustering algorithm is very effective for membership determination. The DBSCAN clustering algorithm can ef- fectively select probable member stars in 3D kinematic space without any assumption about the distribution of the cluster or field stars. Analysis results show that the CMD of member stars is significantly clearer than the one based on 2D kinematics, which al- lows us to better constrain the cluster members and estimate their physical parameters. Using the 472 member stars, the average absolute proper motion and radial velocity are determined to be （PMα, PMδ） = （-2.58 ± 0.22, ＋0.17 ± 0.18） mas yr-1 and Vr = -42.35 ± 0.05 km s-1, respectively. Our values are in good agreement with values derived by other authors.
基金supported by the National Natural Science Foundation of China(Grant No.51276018)
文摘Rotor blades in a radial turbine with nozzle guide vanes typically experience harmonic aerodynamic excitations due to the rotor stator interaction. Dynamic stresses induced by the harmonic excitations can result in high cycle fatigue(HCF) of the blades. A reliable prediction method for forced response issue is essential to avoid the HCF problem. In this work, the forced response mechanisms were investigated based on a fluid structure interaction(FSI) method. Aerodynamic excitations were obtained by three-dimensional unsteady computational fluid dynamics(CFD) simulation with phase shifted periodic boundary conditions. The first two harmonic pressures were determined as the primary components of the excitation and applied to finite element(FE) model to conduct the computational structural dynamics(CSD) simulation. The computed results from the harmonic forced response analysis show good agreement with the predictions of Singh's advanced frequency evaluation(SAFE) diagram. Moreover, the mode superposition method used in FE simulation offers an efficient way to provide quantitative assessments of mode response levels and resonant strength.
