基于时间反转MUSIC的奥氏体不锈钢超声成像分析

奥氏体不锈钢管道广泛应用于核电及某些化工领域等,但奥氏体不锈钢焊缝缺陷的检测仍然比较困难。在时间反转的基础上,采用DORT算法和MUSIC算法对较低信噪比(-5 dB和-10 dB)的缺陷定位进行仿真分析,在定制的奥氏体不锈钢对接焊缝试块上(300 mm×90 mm×30 mm),采用16阵元超声相控阵探头对缺陷进行检测,引入TR-MUSIC算法对缺陷回波信号进行频域分析。结果表明,低信噪比时,TR-MUSIC算法较TR-DORT算法能更好的抑制旁瓣,使信号能量能更好的聚焦于缺陷;在奥氏体不锈钢焊缝相距25 mm的Φ2mm缺陷(深度为50 mm)检测实验数据基础上,TR-MUSIC算法能够更准确定位缺陷。     

工序问题的动态规划算法

提出了一个求解工序问题的动态规划算法,该算法排序含n个工件集合的期望时间为O(n)。     

圆阵波束输出的高分辨方位估计

在研究波束域高分辨ＭＵＳＩＣ算法的基础上,本文把阵元域ＷＳＦ算法推广到波束域中。这两种算法对阵列的形式没有限制,可以应用于水声系统中常见的均匀圆阵波束输出,计算机仿真结果表明,对于圆阵波束输出,波束域ＭＵＳＩＣ算法估计目标方位的性能优于阵元域ＭＵＳＩＣ算法,而波束域ＷＳＦ算法测保持了阵元域ＷＳＦ算法的最佳方位估计性能。     

Multi-competitive interaction of As(III) and As(V) oxyanions with Ca(2+), Mg(2+), PO(3-)(4), and CO(2-)(3) ions on goethite

Complex systems, simulating natural conditions like in groundwater, have rarely been studied, since measuring and in particular, modeling of such systems is very challenging. In this paper, the adsorption of the oxyanions of As(III) and As(V) on goethite has been studied in presence of various inorganic macro-elements (Mg(2+), Ca(2+), PO(3-)(4), CO(2-)(3)). We have used 'single-,' 'dual-,' and 'triple-ion' systems. The presence of Ca(2+) and Mg(2+) has no significant effect on As(III) oxyanion (arsenite) adsorption in the pH range relevant for natural groundwater (pH 5-9). In contrast, both Ca(2+) and Mg(2+) promote the adsorption of PO(3-)(4). A similar (electrostatic) effect is expected for the Ca(2+) and Mg(2+) interaction with As(V) oxyanions (arsenate). Phosphate is a major competitor for arsenate as well as arsenite. Although carbonate may act as competitor for both types of As oxyanions, the presence of significant concentrations of phosphate makes the interaction of (bi)carbonate insignificant. The data have been modeled with the charge distribution (CD) model in combination with the extended Stern model option. In the modeling, independently calculated CD values were used for the oxyanions. The CD values for these complexes have been obtained from a bond valence interpretation of MO/DFT (molecular orbital/density functional theory) optimized geometries. The affinity constants (logK) have been found by calibrating the model on data from 'single-ion' systems. The parameters are used to predict the ion adsorption behavior in the multi-component systems. The thus calibrated model is able to predict successfully the ion concentrations in the mixed 2- and 3-component systems as a function of pH and loading. From a practical perspective, data as well as calculations show the dominance of phosphate in regulating the As concentrations. Arsenite (As(OH)(3)) is often less strongly bound than arsenate (AsO(3-)(4)) but arsenite responses less strongly to changes in the phosphate concentration compared to arsenate, i.e., deltalogc(As(III))/deltalogc(PO(4)) approximately 0.4 and deltalogc(As(V))/deltalogc(PO(4)) approximately 0.9 at pH 7. Therefore, the response of As in a sediment on a change in redox conditions will be variable and will depend on the phosphate concentration level.     

Hartmann－Shack波前传感器的Fourier光学分析

Ｈａｒｔｍａｎｎ－Ｓｈａｃｋ波前传感器是自适应光学中几种重要的波前传感器之一。本文利用Ｆｏｕｒｉｅｒ光学方法分析研究了Ｈａｒｔｍａｎｎ－Ｓｈａｃｋ波前传感器的原理，导出了辐照度分布质心与子孔径波前倾斜的解析表达式，说明了辐照度分布质心及其算法的物理意义，总结了该传感器的若干重要特点。     

基于矢量水听器阵列空域预滤波处理的多重信号分类算法研究

矢量水听器能同时拾取声压和振速信息,在相同的信噪比、阵元数及阵列孔径下,矢量阵定向性能优于声压阵列。目前,以多重信号分类算法(Multiple signal classification,MUSIC)为代表的高分辨定向算法已经广泛应用于矢量水听器阵列中。但是随着信噪比降低、信号源方位间隔减小,传统MUSIC算法定向精度及分辨概率显著下降。本文采用最小二乘法设计适用于矢量水听器水平阵列的矩阵空域滤波器,用于阵列数据的空间滤波预处理,可以对阻带扇面噪声进行有效抑制。由滤波后的数据协方差矩阵可以得到新的噪声子空间,在传统MUSIC算法基础上修正通带扇面内阵列流型的畸变后即可得到滤波后MUSIC算法的方位谱。仿真结果表明,当信噪比较低时,改进算法有效提高了通带扇面内目标方位分辨性能。最后本文对四基元矢量水平阵列海试数据进行了处理,改进算法对窄带信号定向较常规算法-3 dB束宽减小了13°,旁瓣级降低约8 dB。对有一定带宽的行船辐射噪声定向处理得到了更加精确的航迹图,海试数据处理结果证明了该算法的可行性和有效性。     

无界意义下的在线变化分位数回归算法

本文研究了基于核方法下的在线变化损失函数的回归算法.利用迭代和比较原则,得到了算法的收敛速度,并将该结果推广到了更一般的输出空间.     

计算等式约束下多项式在闭长方体上的精确最小值

对于给定的一个实多项式函数f,多项式环R[x1,…,xn]中一个非空的有限子集H以及Rn中一个闭长方体∏n i=1[ai,bi],给出了一个有效算法,用来计算多项式函数f在集合∏n i=1[ai,bi]∩ZeroR(H)上的精确最小值,这里ZeroR为的实零点集。此外,该算法可产生一个最小值点,该点被写成所谓的区间-有理单元表示。相应的有关算法通过Maple软件被编制成一个通用程序,可处理相关实例。