共查询到17条相似文献,搜索用时 62 毫秒
1.
应用化学镀镍的方法实现了氮化铝的金属化。为得到较大的氮化铝金属化层粘附力 ,运用基于稳健估计的神经网络研究氮化铝金属化中化学镀镍的反应参数与金属层粘附力的关系。为使神经网络更加稳健 ,本文根据统计学原理 ,在前馈神经网络基础上 ,采取稳健估计方法改进神经网络。建立了定量预测粘附力性能的模型 ,并进行实验验证。确定金属化工艺中稳定的优化工作区域。结果表明 ,稳健估计方法既有传统神经网络的优点 ,又有较强的抵抗异常值的能力 ,具有较广泛的实用性。 相似文献
2.
应用化学镀镍的方法实现了氮化铝的金属化。为得到较大的氮化铝金属化层粘附力,运用基于稳健估计的神经网络研究氮化铝金属化中化学镀镍的反应参数与金属层粘附力的关系。为使神经网络更加稳健,本文根据统计学原理,在前馈神经网络基础上,采取稳健估计方法改进神经网络。建立了定量预测粘附力性能的模型,并进行实验验证。确定金属化工艺中稳定的优化工作区域。结果表明,稳健估计方法既有传统神经网络的优点,又有较强的抵抗异常 相似文献
3.
通过稀疏重构得到传感器阵列输出数据的稀疏表示模型,研究了单快拍采样情形下的信号到达角(Direction of Arrival, DOA)估计问题。提出了一种基于最小均方误差(Minimum Mean-Square Error, MMSE)准则迭代实现的单快拍到达角估计算法(Iterative Implementation of MMSE, II-MMSE)。该算法将原有的稀疏表示模型中稀疏信号矢量的求解问题,转化为迭代求解稀疏功率对角阵,进而估计多目标信号的DOA。给出了算法的完整实现流程,从理论上分析了II-MMSE算法的迭代收敛性和对阵列模型误差的鲁棒性。仿真结果表明,II-MMSE算法在低信噪比、相干背景、小样本、阵列未校准等条件下都具有良好的测向精度和多目标分辨能力。 相似文献
4.
针对近场源定位问题,提出了一种使用加权L1范数优化进行稀疏信号重构的近场源定位方法。该定位方法分步完成目标的方位和距离估计。为了避免二维优化问题出现,首先利用均匀线阵的对称特性,通过菲涅尔近似,将二维参数估计的近场定位问题转换为类远场阵列的一维参数估计问题,接着将该一维参数估计问题转换为稀疏信号重构问题,通过类MUSIC权向量的构造,使用加权L1范数优化方法重构稀疏空间谱得到目标波达方向;在得到信号波达方向之后,再利用稀疏信号重构的思想求解信号源到阵列的距离。最后,通过数字仿真验证了算法在估计精度和分辨率等方面的优良性能。 相似文献
5.
6.
7.
8.
9.
为了提升稀疏贝叶斯(Sparse Bayesian Learning, SBL)算法在干扰环境下对目标信号的检测能力,提出将频率着色技术(Frequency Coloring, FC)推广至SBL算法中。在SBL-FC算法中,首先将阵列接收信号通过傅里叶变换转换至各个子带,在各子带内利用SBL算法进行波达角估计,输出功率谱。不同于常规的SBL算法仅将各子带的功率谱进行简单地叠加,算法考虑干扰和目标频谱结构的差异性,对各子带进行不同的着色,使得干扰和目标轨迹在方位时间历程图上对应于不同的颜色,从而使得目标轨迹更易被提取。数值仿真和实验数据分析表明,利用目标和干扰频谱结构的差异性可有效提升SBL算法在干扰环境下对目标信号的检测能力。 相似文献
10.
11.
12.
由于机械噪声传播过程中存在反射等多种因素影响,大多数情况下混合噪声分离更适合采用卷积模型,为此提出了一种多频点盲解卷算法。有别于传统的频域盲解卷算法,新算法利用有限的少数几个频率点直接从频域模型恢复出时域噪声信号。算法为瞬时混合盲分离。主成分分析一瞬时混合盲分离结构,首先对给定的每一个频率点执行瞬时混合盲解卷算法,获得噪声源的基本估计,然后再经过主成分分析和第二次盲源分离。提高分离性能和增加算法鲁棒性。由于算法不需要对所有频率点执行瞬时混合分离,计算量小,同时也不存在传统频域盲解卷算法排列顺序不确定性的缺点,具有较好的应用价值。仿真实验证实了新算法能有效地分离机械噪声信号。 相似文献
13.
动态转速的高精度测量是高性能运动控制中的重要问题,尤其对于低速运动情况.针对该问题,提出一种新型实时转速估计方法,对系统外部扰动和建模误差,采用扰动观测器进行估计,再将其输出引入转速观测器,从而保证转速估计的收敛性,并增强其鲁棒性.以直流高精度伺服系统为例,进行了仿真和实验研究,结果表明该方法有效,鲁棒性强,且对角位置量化噪声不敏感. 相似文献
14.
Studies on the Effects of Estimator Selection in Robust Parameter Design under Asymmetric Conditions
Gregory L. Boylan Byung Rae Cho 《Quality and Reliability Engineering International》2013,29(4):571-582
The primary goal of robust parameter design (RPD) is to determine the optimum operating conditions that achieve process performance targets while minimizing variability in the results. To achieve this goal, typical approaches to RPD problems use ordinary least squares methods to obtain response functions for the mean and variance by assuming that the experimental data follow a normal distribution and are relatively free of contaminants or outliers. Consequently, the most common estimators used in the initial tier of estimation are the sample mean and sample variance, as they are very good estimators when these assumptions hold. However, it is often the case that such assumed conditions do not exist in practice; notably, that inherent asymmetry pervades system outputs. If unaccounted for, such conditions can affect results tremendously by causing the quality of the estimates obtained using the sample mean and standard deviation to deteriorate. Focusing on asymmetric conditions, this paper examines several highly efficient estimators as alternatives to the sample mean and standard deviation. We then incorporate these estimators into RPD modeling and optimization approaches to ascertain which estimators tend to yield better solutions when skewness exists. Monte Carlo simulation and numerical studies are used to substantiate and compare the performance of the proposed methods with the traditional approach. Copyright © 2012 John Wiley & Sons, Ltd. 相似文献
15.
16.
John A. Simmons 《Journal of research of the National Institute of Standards and Technology》1991,96(3):345-369
A new technique, root projection (RP), is given for quantitative deconvolution of
causal time series in the presence of moderate amounts of noise. Deconvolution is
treated as a well-conditioned but underdetermined problem and a
priori information is employed to obtain comparable noise reduction to
that achieved by singular value decomposition (SVD) techniques while providing more
accurate frequency information about the inverse. Two detailed examples arc given.
The first gives noise analysis for alternate methods for deconvolution with a
Gaussian kernel. The second example presents a model acoustic emission transducer
calibration problem with typical noisy and incomplete output data. This example is
treated by the use of a robust cross-cutting algorithm combining both the RP and SVD
methods. 相似文献
17.
Qing‐Hua Lu Xian‐Min Zhang 《International journal of imaging systems and technology》2007,17(6):333-340
Gradient‐based techniques represent a very popular class of approaches to estimate motions. A robust multiscale algorithm of hierarchical estimation for gradient‐based motion estimation is proposed in this article using a combination of robust statistical method and multiscale technique. In such a multiscale approach of hierarchical estimation, motion at each level of the pyramid is estimated using different gradient filters. The iterative multiscale estimation begins by using five‐tap central filter, and it is switched to nine‐tap Timoner filter after a few iterations. In addition, robust M‐estimators are applied at each level of the pyramid to overcome the problem of the outliers caused by illumination variations and motion discontinuities in motion estimation. Experimental simulations show that the new algorithm not only provides an improvement in estimator accuracy, but also achieves computational speedups. © 2008 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 17, 333–340, 2007 相似文献