稀疏图的k-frugal列表染色

图G的k-frugal列表色数一般记作chkG，关于稀疏的k-frugal列表色数上界，有以下3个结论：? k≥3，如果图G满足madG<3-a（其中0<a≤13）且ΔG≥k+3a-3，则chkG=?k-1+1；? k≥4，如果图G满足madG<3，则chkG≤?k-1+2；? k≥4，如果图G满足madG<103，则chkG≤?k-1+3。     

分级加载与波浪荷载耦合作用下软黏土动力响应特性

以长江口长兴岛潜堤后方滩涂圈围工程为研究背景，采集堤坝地基土样，根据实际工况应力路径设计不同围压和动静应力组合，对淤泥质软黏土进行动三轴试验，探究地基土在分级施工加载与波浪荷载耦合作用下软黏土应变、孔压特性以及强度弱化特性。研究结果表明：耦合应力加载后土样的应力-应变曲线呈现应变软化和强度弱化特性，分级加载效应可使应变及孔压增量减少，且应变增量、孔压增量与循环振后静强度折减度，均和固结围压、静偏应力比、动偏应力比等因素相关。当动偏应力比相同时，深度较浅土层围压较小，静偏应力比反而较大，其应变与孔压增量越大，静强度衰减越明显。随着动偏应力比增大，土单元的应变增量与孔压增量、强度折减度有不同程度的增大，并且不同深度的差异性较大。在此基础上，分别对强度折减度β、广义综合剪应变γgs及广义综合孔压增量比Δu/σ3进行了量化，并对（γgs-β）及（Δu/σ3-β）相关性进行了探讨。研究结果对于指导具体工程实践具有一定的参考价值。     

基于圆柱绕流的气动声源识别方法

以气流流经固体壁面产生的气动声源为研究对象，以力点源对应的偶极子气动声源声波动方程的声压解为基础，利用流场中声源辐射声压和脉动力、脉动力和压力梯度的关系，建立偶极子气动声源辐射声压与流场压力梯度的关系式；利用两偶极子声源可组成一四极子声源的概念，建立四极子气动声源辐射声压与流场脉动速度的关系式。最后，以圆柱绕流为研究对象，采用上述两关系式并结合数值仿真计算方法，得到圆柱绕流的偶极子和四极子气动声源大小和分布特征。结果表明：偶极子气动声源向远场的声辐射声压由??p?t决定，四极子气动声源向远场的声辐射声压由?2ur2?t2决定。     

求解线性方程组稀疏解的稀疏贪婪随机Kaczmarz算法

求解线性方程组的稀疏解在图像重构、信号处理和机器学习等领域中具有广泛的应用，通过引入l1-范数正则化，可以转化为求解一个约束优化问题。基于一种选择系数矩阵工作行的概率准则，提出了稀疏贪婪随机Kaczmarz算法，并给出了有噪声干扰和无噪声干扰情况下该算法的收敛性分析。理论表明本文算法的收缩因子小于随机稀疏Kaczmarz算法的收缩因子。数值实验验证了本文算法的有效性。     

矩形截面高层建筑立面上面风压极值的尺度折减系数

通过一系列刚性模型测压风洞试验，研究了矩形截面高层建筑立面面风压极值的变化特征，并与我国建筑结构荷载规范GB 50009—2012中的相关规定进行了对比，分析了受压面的面积、水平尺度及竖向尺度的影响，给出了更合理的尺度参数。试验结果表明，随受压面尺度的增加，面风压极值逐渐减小，当足尺面积达到50 m²时，面风压正极值的尺度折减系数在0.9左右，而负极值的尺度折减系数在0.8左右。我国建筑结构荷载规范对围护结构风荷载的尺度折减方法将导致面风压极值的低估，尤其是面风压正极值。以作用面面积、对角线长度和宽度作为尺度参数描述面风压极值折减系数的变化规律都不合理，建议采用以受压面的水平尺度b、竖向尺度h和整个建筑迎风面宽度B确定的综合尺度参数（b0.85h0.15/B）来描述建筑立面上面风压极值的尺度折减系数的变化规律。     

路径直积图的意大利控制数

研究了路径直积图P_n×P_m的意大利控制数。结合计算机构造证明和数学推导证明，确定了P_n×P₁、P_n×P₂和P_n×P₃的意大利控制数，并给出了P_n×P_m （m≥4）意大利控制数的界。     

加湿除湿技术用于高盐工业废水脱盐的实验研究

通过加湿除湿蒸发实验，对高盐工业废水34个不同水样采用了加湿除湿技术进行脱盐处理。在所得实验数据的基础上，对加湿除湿技术在实际工业中的普遍应用进行了可行性研究，并结合不同高盐工业废水的水质特点和处理效果，总结了判定该方法处理高盐工业废水的效果的一般规律。结果表明，实验中所有废水的冷凝液的电导率下降百分比ησ均在80.0%以上，最高可达99.9%；冷凝液的盐度均可达到生化处理标准，其中，有32个废水样的冷凝液的电导率可降到2 000 μS?cm^-1以下。另外，加湿除湿技术对沸点比水高的盐类物质的去除效果较为彻底，对沸点比水低且电离常数较大的盐类物质也有较好的去除效果，而对沸点比水低且电离常数较小的盐类物质的去除效果较差。     

基于卷积神经网络-长短期记忆的施工期盾构管片上浮过程预测模型

为了实现施工期盾构管片上浮过程的智能预测，采用动力水准仪对施工期盾构管片上浮过程进行自动化监测并建立了基于卷积神经网络?长短期记忆（CNN-LSTM）深度学习算法的管片上浮过程智能预测模型。结果表明：管片上浮阶段呈现出“阶梯状”，即管片上浮主要发生在盾构掘进期间，且掘进状态的上浮量最大，占峰值的75.24 %~98.29 %；CNN-LSTM模型对施工期盾构管片上浮过程具有较好的预测效果，在训练集上的均方误差MSE、平均绝对误差MAE和决定系数R2分别为0.038 7、0.148 2和0.999 3，在测试集上为0.030 7、0.138 9和0.801 9；相较于反向传播（BP）模型，CNN-LSTM模型在训练集与测试集上的性能均有所提升，且测试集的提升更明显，最高可达89.71 %。研究结果可为盾构管片上浮的现场实测及预防处治提供新思路。     

求解大型线性最小二乘问题的贪婪Gauss-Seidel方法

基于一种选择系数矩阵A的工作列的策略，提出了求解大型线性最小二乘问题的一种不同的贪婪Gauss-Seidel方法，并对该方法进行了收敛性分析。数值实验表明，在相同的精度下，所提方法在计算时间上优于文献提出的贪婪随机坐标下降方法。     

高密度路网区高速公路可变信息标志版面要素组合

为确定适用于高密度路网区高速公路的可变信息标志（VMS）版面要素组合，综合运用小型驾驶模拟器和眼动仪采集了20名驾驶员视认4种VMS版面和2种普通标志的数据，采用指标方差分析检验了6种版面诱导结果、驾驶操纵稳定性、驾驶操纵负荷、视认行为的差异性，结果表明，VMS版面4与普通标志牌无显著差异，不会对行车安全造成影响。采用指标因子分析法提出综合反映驾驶员各项驾驶行为指标特征的参数S，实现对驾驶行为的多元量化评价，结果表明VMS版面4最佳。综合以上分析，推荐高密度路网区高速公路VMS版面要素组合形式为，路网形状采用拓扑图表示，高速公路用平纵相连的光带表示，其他道路线形用短横线简略表示，互通用圆形表示；红、黄、绿色分别表示拥堵、缓行、畅通3种路况；各互通处采用白色三角标注路段指向，枢纽、一般互通处分别标注地名和互通名，并对驾驶员所处位置进行醒目标注。     

厄尔尼诺-南方涛动影响下月尺度水文模型参数可移植性

以浙江省3个流域作为研究对象，采用两参数月水量平衡模型（XM模型）和四参数月水量平衡模型（ABCD模型），结合极大似然不确定性估计方法（GLUE）对水文模拟的不确定性区间进行估计，在此基础上，研究不同模型在ENSO（厄尔尼诺-南方涛动）和不同流域地理情况下的参数移植能力。结果表明，厄尔尼诺时期的参数移植至拉尼娜时期，模型模拟效果要略好于拉尼娜时期的参数移植至厄尔尼诺时期；参数少的模型（XM模型）参数移植合格率要优于参数多的模型（ABCD模型）；相邻流域间的参数移植效果要明显优于距离较远的流域；ABCD模型具有更好的空间可参数移植能力。     

高腐蚀条件下用铝合金材料腐蚀机理

铝合金具有低密度、低熔点、高比强度及优良的耐腐蚀性能等特点，被广泛用于航空航天、建筑、船舶等领域。在服役过程中，铝合金的表层氧化膜易受到环境中活性阴离子的破坏而发生腐蚀，对其性能造成严重的损害，故研究铝合金在高腐蚀性环境的腐蚀行为对工程选材具有非常重要的指导意义。选择6061铝合金、2195铝锂合金和7075铝合金为研究对象，对其在特定腐蚀介质中的腐蚀过程和力学性能进行分析，研究了铝合金在特定腐蚀介质中腐蚀形貌与力学性能的变化规律。结果表明：腐蚀初期，在高Cl^-、NO₂^3-、SO₂^4-离子浓度的腐蚀环境中，3种铝合金的氧化膜受到阴离子破坏后发生点腐蚀，使基体暴露在腐蚀环境中，进而发生电化学腐蚀，6061铝合金和2195铝锂合金腐蚀方式是由点腐蚀向面腐蚀转变，7075铝合金腐蚀方式为晶间腐蚀；经过腐蚀后6061铝合金能保持稳定的强度和塑性，7075铝合金和2195铝锂合金的强度和塑性都明显降低。     

CoPt-Co hybrid supported on amino modified SiO_2 nanospheres as a high performance catalyst for hydrogen generation from ammonia borane

CoPt-Co hybrids were successfully supported on amino modified SiO_2 nanospheres by a chemical reduction method at a temperature of 278 K.The solid carrier i.e.amino modified SiO_2,provides numerous anchoring sites for the metal nanoparticles(NPs)to improve the dispersion while reducing the size of metal NPs.The supported NPs displayed a narrow particle size distribution on the SiO_2 surface with an average diameter of 12 nm.The XRD results alongside with the binary alloy phase diagram suggest that the resulted NPs are bimetallic,composed of CoPt and amorphous Co.Among the prepared materials,the solid with the specific composition of SiO_2@Pt_(0.1)Co_(0.9)was proved to be effective catalyst for ammonia borane(AB)hydrolysis in aqueous solution.The turnover frequency(TOF)value of the supported nanocatalyst was 25.59mol_(H_2)min~(-1)·mol_M~(-1),almost twice as that of unsupported Pt_(0.1)Co_(0.9)NPs while the activation energy was 37.05 kJ mol~(-1).Furthermore,the SiO_2@Pt_(0.1)Co_(0.9)composite manifested high catalytic activity even after five cycles of reuse.     

Microstructure/texture evolution maps to optimize hot deformation process of near-α titanium alloy

The microstructure/texture evolution(MTE,for short) map and processing map of a new near a titanium alloy Ti65 were constructed in order to investigate the workability and microstructure evolution of hot deformation.The processing map illustrated four domains,two summit domains and two instability domains.The morphologies of the a phase changed from the spheroidization(α+β region) to the deformed and elongated β grains(near the βtransus temperature T_β),and then to the obvious dynamic recrystallization(DRX)(β region) with the temperature rising from 930 ℃ to 1140℃.Deformation in the α+β field mainly generated the texture component with [0001]or [0223] parallel to radial directions(RDs).While deformation in the β phase field formed two types of texture component with [0001] parallel to RDs and [2110] parallel to compression direction.An optimized processing map was summarized by overlaying the macro-instability map on the original processing map,and the instability domain of Ti65 alloy was confirmed in the area with the strain rate higher than 0.01 s~(-1).     

Microstructure evolution and mechanical properties of Ti-8Nb-2Fe-0.2O alloy with high elastic admissible strain for orthopedic implant applications

A Ti-8Nb-2Fe-0.2O(wt.%) alloy with high strength,high elastic admissible strain(δ) and low cost was designed using d-electron theory combined with electron-to-atom ratio(e/a) approach.Interstitial oxygen was introduced to strengthen the matrix of the alloy.The β-solution treated alloy was mainly composed of α " martensite with internal {111}_(α") type Ⅰ nano-twins.The α " martensite with hexagonal-like crystal structure caused by interstitial oxygen showed a high strength of 1.1 GPa but limited ductility.The alloy generated equiaxed fine-grained a phase embedded by β matrix via hot rolling and subsequent annealing in α+β phase field.The obtained alloy indicated a good combination of mechanical properties with ultimate tensile strength,Young's modulus,ductility and δ value of 1029 MPa,74 GPa,21% tensile elongation and 1.32%,respectively.These findings demonstrate that interstitial oxygen and martensitic nano-twins can be used to strengthen the soft α" martensite and high elastic admissible strain can be obtained by formation of equiaxed fine-grained α phase embedded by βmatrix in this Ti-8Nb-2Fe-0.2O alloy for orthopedic implant.     

Earth-abundant photovoltaic semiconductor NaSbS_2 in the rocksalt-derived structure:A first-principles study

NaSbS_2 was recently proposed as a novel photovoltaic semiconductor with earth-abundant component elements,but its fundamental material properties have not been well studied.The systematical first-principles calculations for its electronic,optical and defect properties were carried out in the present study,and the results show that:i)NaSbS_2 in the rocksalt-derived structure has a quasi-direct band gap and thus may have long minority carrier lifetime;ii) its absorption coefficients are as high as 10~4～10~5 cm~(-1) for the visible light and almost isotropic despite that the structure is distorted relative to the high-symmetry rocksalt structure;iii) the effective masses of the electron and hole carriers are anisotropic with much larger values along the z direction than in the x-y plane,and hence the orientational control of thin films should be important for enhancing the photovoltaic performance;iv) the valence and conduction band edges of NaSbS_2 are close to those of CuGaSe_2) so the n-CdS/pCuGaSe_2 device structure can be inherited to form the n-CdS/p-NaSbS_2 solar cells;v) the acceptor defects (Na_(Sb)antisites and Na vacancies) have very high concentration,making the synthesized NaSbS_2 always be p-type;vi)the S-rich condition can suppress the formation of deep-level donor defects (S vacancies and Sb_(Na) antisites) and therefore should be adopted for fabricating high-efficiency NaSbS_2 solar cells.     

Modeling the hot working behavior of near-α tita nium alloy IMI 834

The hot working behavior of near-a titanium alloy IMI 834 with a duplex starting microstructure was studied using the technique of processing map.The processing map was interpreted in terms of the microstrucmral processes occurring during deformation,based on the values of dimensionless parameter η which represents the energy dissipation through microstructural processes.An instability criterion（ξ<0） was also applied to demarcate the flow instability regions in the processing map.Both the parameters（η and ξ） were computed using the experimental data generated by carrying out hot compression tests over a range of temperatures（850-1060℃） and strain rates(3×10^-4-1/s).The deterministic domains observed under the investigated temperature and strain rate conditions were attributed to continuous dynamic recrystallization or globularisation of a lamellae,dynamic recrystallization and growth of β grains through microstructural observations.An unified strain compensated constitutive equation was established to describe the hot working behavior of the material in the selected temperature-strain rate range.The established constitutive equation was validated using standard statistical parameters such as correlation coefficient and average absolute relative error.