复杂目标的FDTD几何-电磁建模方法

本文提出一种利用DirectX建立复杂目标FDTD模型的实用方法.运用DirectX分析目标模型的整体几何特性和材质特性,从不同侧面作多束穿透模型的射线,获得目标几何结构细节,判断剖分网格是否在模型内部,根据材质信息得到相应位置的电磁参数,最后按照FDTD计算要求建立复杂目标的空间电磁信息文件.文章建立了金属球体的FDTD模型并计算了它的电磁散射特性,用所得结果验证建模方法的有效性,进而给出了BLITZ卡车、M60坦克的FDTD模型.     

基于复数帧段特征的语音情感识别方法

在现有的电磁场算法中,时域有限差分法(FDTD)的使用越来越广泛。但是当解决曲面问题时,原有的FDTD方法就不太适用了。而且,大多数的电气设备为多重材料结构和不连续的曲线表面,传统的FDTD算法会产生发散和不稳定现象。由此提出了一种基于非标准曲线算子的三维高阶FDTD算法,用于解决复杂介质结构电气设备的电磁兼容性问题。根据代数拓扑学,将各种不规则形状的网格转化为合适的二重网格用公式表示出来,在优化整体性能的同时降低色散误差。通过计算结果可以看出,高阶非标准FDTD算法的准确度极高,而且没有任何色散情况出现,并且高阶非标准FDTD算法还可以节省大约85%的CPU和内存使用。证明了该FDTD算法能有效提高计算精度并减少计算资源的消耗。     

加载天线目标RCS的FDTD计算与分析

提出一种带角点的细导线模型和传统FDTD相结合的方法分析加载斜拉式线天线目标散射特性.为进一步减小目标离散时的台阶化误差和节省计算资源(内存和时间),给出一种简化的共形FDTD(SCFDTD)算法.通过坐标旋转使天线长段平行于某固定坐标轴,然后建立带有角点的细导线FDTD模型.以变形网格面积与正常网格面积之比1/6作为SCFDTD算法的判据,既简化了已有CFDTD算法的复杂度,又不失其物理意义.与传统FDTD方法和实验测量结果对比数值计算例子表明了该方法的可行性与有效性.     

改进共形FDTD算法及其在RCS计算中的应用

本文首先提出一种结合了三角近似与改进局部网格的共形FDTD算法,然后将该算法应用于金属目标的雷达散射截面计算。实验结果表明,该算法简单高效,不需要通过减小时间步长的方式就能得到较高精度的稳定解。     

碰撞等离子体的高阶FDTD算法

给出了电磁波在均匀、碰撞等离子体中传播的四阶时间和四阶空间FDTD算法.该算法比Yee氏FDTD算法每一个网格每一维增加一个存储单元,与常规的二阶等离子体FDTD算法相同.由于采用四阶时间和四阶空间近似,因此该算法能有效地减小数字色散误差,其频带宽度比二阶算法的频带宽度更宽.为了验证该高阶算法的正确性,对均匀、碰撞等离子体平板的电磁波反射系数进行了计算,并与解析结果、二阶FDTD计算结果进行了比较,证明了该算法的高效和精确.     

一种改进的FDTD网格剖分算法

分析了时域有限差分(FDTD)网格的生成原理,提出了一种新型非均匀FDTD网格生成算法.该算法通过读取模型获得轴线上的不连续分界点,将整个空间沿x轴、y轴和z轴方向各自分成多个区间;然后通过各区间的长度及其在模型中所处的位置,来确定该不同区间剖分时所采用的具体网格生成算法,由此得到一种新型的分区递变非均匀网格.此算法能够完全由程序实现,避免了普通网格剖分算法所带来的人为误差.在最大限度降低网格数量的同时,克服了剖分算法中由于最大网格过大而导致的高频数值色散问题,使得最后的FDTD程序具有计算时间短、收敛速度快的优点.     

多时间步长时域有限差分法分析微波电路

提出在非均匀网格空间的粗网格区域采用大时间步长,细网格区域采用小时间步长的多时间步长时域有限差分(MTS -FDTD)法,并对其稳定性条件进行了分析。该方法通过应用多个时间步长替代传统非均匀时域有限差分(NU- FDTD)法的单一时间步长的算法来提高计算速度。通过应用MTS- FDTD方法模拟自由空间中给定激励下某点的场和微带低通滤波电路。数值结果表明,与传统NU- FDTD法相比,MTS- FDTD法在保持同等计算精度的条件下计算速度提高了46%以上。     

基于FDTD算法的新型亚网格技术

提出了一种基于空间滤波（Spatial Filtering-Finite-Difference Time-Domain,SF-FDTD）算法的亚网格技术,使得FDTD算法的Courant-Friedrich-Levy（CFL）稳定性条件可通过空间频域滤波操作得以提高,从而获得高稳定度FDTD算法.进一步将SF-FDTD算法应用到亚网格技术中,可使亚网格区域时间步长的选取取与粗网格一致,从而极大地提高了计算效率.数值计算结果表明,在求解带有精细结构的电磁问题上,所提算法具有较高的准确性和有效性.     

三维情况下平面介质交界面处二阶精度FDTD方程

提出了一种新的二阶精度时域有限差分法(FDTD)算法,针对研究区域介质填充不均匀,包含平面介质交界面的情况.新算法在交界面处利用非均匀网格建模,通过积分形式Maxwell方程的离散和交界面处连续场分量的泰勒级数展开,给出了平面介质交界面处三维情况下的二阶精度FDTD方程.该方法在保证介质交界面处电磁场量二阶精度的同时,合理划分不同介质区域粗细网格,在不增加计算容量和计算时间的基础上,有效地提高算法整体的计算精度.最后对介质谐振腔结构和微带电路进行模拟,验证了本文算法的精度高于标准的FDTD方法.     

用于电磁波多尺度问题求解的高效FDTD-PITD混合算法北大核心CSCD

针对电磁波多尺度问题的高效仿真需求,提出了基于亚网格技术的时域有限差分(FDTD)方法与时域精细积分(PITD)方法的混合数值算法。该混合算法的基本思想是采用局部亚网格技术分别对精细结构区域以及其他区域进行剖分,并应用FDTD方法和PITD方法分别对粗网格区域与细网格区域进行求解,同时构建信息交互策略交换细网格区域与粗网格区域的计算信息。一方面该方法减少了电磁波多尺度问题的网格剖分数目,显著降低了内存需求;另一方面由于应用于细网格区域的PITD方法不受Courant-Friedrich-Levy(CFL)数值稳定性条件的限制,该混合方法能够采用较大的时间步长进行仿真,减少了迭代步数以及CPU执行时间。数值计算结果验证了混合算法的稳定性、可行性以及高效性。     

FDTD方法中散射体曲面边界的网格拟合

在ＦＤＴＤ方法中为了精确拟合散射体曲面边界，本文提出了分区域剖分方法，以便提高计算精度，减少网格数目，节省计算机资源和降低ＦＤＴＤ数值色散，金属三角形柱表面电流的计算与实验结果的比较，验证了此方法的有效性。在直角坐标系下，研究了ＦＤＴＤ法ＣＰ剖分技术的计算精度，将ＣＰ剖分技术推广应用于涂敷金属目标的散射问题，由Ｍａｘｗｅｌｌ方程的积分形式推导出了在涂敷吸波介质厚度很薄的情况下，散射体表面附近网格上场量迭代方程，给出了二维涂敷和部分涂敷金属目标ＲＣＳ计算结果。     

Three-Dimensionally Nonorthogonal Alternating-Direction Implicit Finite-Difference Time-Domain Algorithm for the Full-Wave Analysis of Microwave Monolithic Circuit Devices

An alternating-direction implicit finite-difference time-domain (FDTD) algorithm is applied to the full wave analysis of microwave integrated circuit devices. A 3-D multidomain method is developed in nonorthogonal coordinates. Nonorthogonal grids are only used for the anomalistic regions of a complex structure, whereas the standard FDTD lattice is used for the other regions. By using the Jacobian coordinate transformation, curvilinear coordinates can be converted into conventional FDTD format expediently. The perfectly matched layer is used to truncate the boundary. Accurate griddings using the new scheme are obtained, and the complexity of the algorithm is minimal. To illustrate the theory, a sinusoidal plane wave and a Gaussian pulse that propagate through a localized nonorthogonal grid space is used, and the stability of our code is examined. A newly developed compact microstrip bandpass filter is analyzed using the proposed method. The simulated results agree very well with measurements. As compared to other nonorthogonal FDTD (NFDTD) method, the proposed algorithm is much more efficient than other NFDTD counterpart when complex structures are analyzed.      

FDTD local grid with material traverse

Often, a finite-difference time-domain (FDTD) calculation requires a relatively higher mesh resolution in only small subvolumes of the total mesh space. By locally applying finer grids (local grids) to these volumes, the necessary resolution can be obtained. Computation time and memory requirements may be far less than for an FDTD space with the smaller mesh resolution throughout. In many situations, it is important that these local-grids function when materials traverse the main-grid-local-grid (MG-LG) boundary surfaces, since the volumes that require local grids may not be isolated in a homogeneous medium. A local-grid method, which allows dielectric and/or conducting materials to traverse the boundaries, is developed. Three different FDTD problems that utilize the local-grid method are used as validation tests. Results are compared with uniform mesh FDTD solutions     

CFDTD法在涂敷金属目标TM散射问题中的应用研究

本文对ＦＤＴＤ方法中的ＣＰ剖分析技术在二维情况下进行了详细的研究，推导出并数值实现了涂静金属导体附近网格点上场量的迭代方程。在二维ＴＭ情况下，对涂敷或半涂敷金属柱体的远区ＲＣＳ及近场分布进行计算，根据计算实例解析结果考查了ＣＦＤＴＤ的计算精度。     

同轴馈电耦合微带贴片天线的时域有限差分法分析

提出一种融合曲线坐标和直角坐标的区域网格划分方法。在同轴线区域采用曲线坐标网格,用以代替传统的阶梯剖分,在其它区域及边界上采用直角坐标网格,从而极大地提高了计算效率。同时,在时域有限差分法中还采用了非均匀网格划分,PML吸收边界和Prony外推法等技术,对同轴馈电的耦合微带贴片天线进行了分析。     

机载短波天线间隔离度的全波分析

对飞机等复杂形体及机载短波天线进行全波ＦＤＴＤ数值模拟所需的数学建模方法的介绍,研究了机载短波天线间隔离度的分析计算方法,并以波音７０７－３２０Ｂ飞机为例计算了机上４付短波天线间隔离度。所用方法及所得结果可作为飞机电磁兼容分析与设计的手段和依据。     

Three-dimensional simulation of eccentric LWD tool response in boreholes through dipping formations

We simulate the response of logging-while-drilling (LWD) tools in complex thee-dimensional (3-D) borehole environments using a finite-difference time-domain (FDTD) scheme in cylindrical coordinates. Several techniques are applied to the FDTD algorithm to improve the computational efficiency and the modeling accuracy of more arbitrary geometries/media in well-logging problems: (1) a 3-D FDTD cylindrical grid to avoid staircasing discretization errors in the transmitter, receiver, and mandrel geometries; (2) an anisotropic-medium (unsplit) perfectly matched layer (PML) absorbing boundary condition in cylindrical coordinates is applied to the FDTD algorithm, leading to more compact grids and reduced memory requirements; (3) a simple and efficient algorithm is employed to extract frequency-domain data (phase and amplitude) from early-time FDTD data; (4) permittivity scaling is applied to overcome the Courant limit of FDTD and allow faster simulations of lower frequency tool; and (5) two locally conformal FDTD (LC-FDTD) techniques are applied to better simulate the response of logging tools in eccentric boreholes. We validate the FDTD results against the numerical mode matching method for problems where the latter is applicable, and against pseudoanalytical results for eccentric borehole problems. The comparisons show very good agreement. Results from 3-D borehole problems involving eccentric tools and dipping beds simultaneously are also included to demonstrate the robustness of the method.     

基于面元模型的FDTD网格及涂敷层网格生成技术

提出了一种基于面元模型的FDTD网格剖分技术和涂敷层网格的生成技术。这种网格剖分技术可以方便地剖分封闭的面元模型和实体模型。针对均匀涂敷复杂模型难以直接产生的问题,提出在FDTD网格模型基础上,生成涂敷网格模型的方法。这种方法很容易得到任意复杂FDTD网格模型的涂敷层网格模型,克服了复杂表面模型和实体模型直接产生涂敷模型的困难。     

Accurate and efficient GPU ray‐casting algorithm for volume rendering of unstructured grid data

We present a novel GPU‐based ray‐casting algorithm for volume rendering of unstructured grid data. Our volume rendering system uses a ray‐casting method that guarantees accurate rendering results. We also employ the per‐pixel intersection list concept in the Bunyk algorithm to guarantee an accurate result for non‐convex meshes. For efficient memory access for the lists on the GPU, we represent the intersection lists for all faces as an array with our novel construction algorithm. With the intersection lists, we perform ray‐casting on a GPU, and a GPU thread handles each ray. To increase ray‐coherency in a thread block and improve memory access efficiency, we extend a prior image‐tile‐based work distribution method to fit modern GPU architectures. We also show that a prior approach using a per‐thread local buffer to reduce redundant computation is not appropriate for modern GPU architectures. Instead, we take an on‐demand calculation strategy that achieves better performance even though it allows duplicate computations. We applied our method to three unstructured grid datasets with different characteristics. With a GPU, our method achieved up to 36.5 times higher performance for the ray‐casting process and 19.7 times higher performance for the whole volume rendering process compared with the Bunyk algorithm using a CPU core. Also, our approach showed up to 8.2 times higher performance than a GPU‐based cell projection method while generating more accurate rendering results. These results demonstrate the efficiency and accuracy of our method.