NEURAL MODELS FOR THE BROADSIDE-COUPLED V-SHAPED MICROSHIELD COPLANAR WAVEGUIDES

This article presents a new approach based on multilayered perceptron neural networks (MLPNNs) to calculate the odd-and even-mode characteristic impedances and effective permittivities of the broadside-coupled V-shaped microshield coplanar waveguides (BC-VSMCPWs). Six learning algorithms, bayesian regulation (BR), Levenberg-Marquardt (LM), quasi-Newton (QN), scaled conjugate gradient (SCG), resilient propagation (RP), and conjugate gradient of Fletcher-Powell (CGF), are used to train the MLPNNs. The neural results are in very good agreement with the results reported elsewhere. When the performances of neural models are compared with each other, the best and worst results are obtained from the MLPNNs trained by the BR and CGF algorithms, respectively.     

Control and diagnostic of the complex impedance of selected perovskite compounds

In this work the diagnosis control of the complex impedance of selected perovskite compounds versus artificial neural network model optimized with the Levenberg–Marquardt algorithm is performed as detection of aging and degradation of materials usually requires destructive testing. The artificial neural network optimized by the Levenberg–Marquardt algorithm used in this work allows us to monitor the materials (LaNd) SrMnCrO₃ in a non-destructive manner. This intelligent control is done by calculating the complex impedance which reveals reliable information on the phenomenon of transport in materials. The method overcomes the problem of the lack of a mathematical expression between the input parameters (temperature, doping, and frequency) and the necessary parameters for computing the impedance (bulk resistance, grain boundary resistance, and the two parameters of the constant phase element impedance A₀ and P). The robustness and performance of the artificial neural network model was verified by introducing additional noise and by using the root mean square error and the R-square.     

Chemometrics and its applications to x‐ray spectrometry

Recent developments in curve fitting, multivariate calibration, and pattern recognition in chemometrics, and their application to x‐ray spectrometry, are reviewed. Relatively innovated algorithms, namely genetic algorithms, neural networks and support vector machines, are discussed. Together with the three algorithms, the performances of different algorithms are compared briefly, which mainly includes principal component analysis, partial least‐squares regression, factor analysis, cluster analysis, nearest neighbor methods, linear discriminant analysis, linear learning machine, and soft independent modeling of class analogy. In general, the chemometrics methods are superior to the conventional methods, such as Fourier transform and Marquardt–Levenberg algorithms, to a certain extent. Copyright © 2006 John Wiley & Sons, Ltd.     

Noise-robust realization of Turing-complete cellular automata by using neural networks with pattern representation

A modularly-structured neural network model is considered. Each module, which we call a ‘cell’, consists of two parts: a Hopfield neural network model and a multilayered perceptron. An array of such cells is used to simulate the Rule 110 cellular automaton with high accuracy even when all the units of neural networks are replaced by stochastic binary ones. We also find that noise not only degrades but also facilitates computation if the outputs of multilayered perceptrons are below the threshold required to update the states of the cells, which is a stochastic resonance in computation.     

Perceptron-like computation based on biologically-inspired neurons with heterosynaptic mechanisms

Perceptrons are one of the fundamental paradigms in artificial neural networks and a keyprocessing scheme in supervised classification tasks. However, the algorithm they provideis given in terms of unrealistically simple processing units and connections andtherefore, its implementation in real neural networks is hard to be fulfilled. In thiswork, we present a neural circuit able to perform perceptron’s computation based onrealistic models of neurons and synapses. The model uses Wang-Buzsáki neurons withcoupling provided by axodendritic and axoaxonic synapses (heterosynapsis). The maincharacteristics of the feedforward perceptron operation are conserved, which allows tocombine both approaches: whereas the classical artificial system can be used to learn aparticular problem, its solution can be directly implemented in this neural circuit. As aresult, we propose a biologically-inspired system able to work appropriately in a widerange of frequencies and system parameters, while keeping robust to noise and error.     

Superpositional Quantum Network Topologies

We introduce superposition-based quantum networks composed of (i) the classical perceptron model of multilayered, feedforward neural networks and (ii) the algebraic model of evolving reticular quantum structures as described in quantum gravity. The main feature of this model is moving from particular neural topologies to a quantum metastructure which embodies many differing topological patterns. Using quantum parallelism, training is possible on superpositions of different network topologies. As a result, not only classical transition functions, but also topology becomes a subject of training. The main feature of our model is that particular neural networks, with different topologies, are quantum states. We consider high-dimensionaldissipative quantum structures as candidates for implementation of the model.     

自适应光学系统变形镜控制电压预测

在校正大气湍流畸变波前相差的自适应光学系统中,利用基于Levenberg-Marquardt学习算法的非线性反向传播神经网络技术（LMBP）对变形镜控制电压进行预测。以对受横向风影响的大气湍流畸变波前斜率数据为研究对象,通过数值仿真方法,研究了基于LMBP算法的自适应光学系统变形镜电压非线性预测控制算法。通过实验发现,预测电压和变形镜实际控制电压拟合效果良好。讨论了回溯帧数对预测效果的影响,并与基于递推最小二乘（RLS）算法的线性预测算法进行比较。对比结果表明,基于LMBP算法的非线性电压预测方法比基于递推最小二乘法的线性电压预测方法能更有效地降低系统由伺服延迟引起的误差。     

北京谱仪III上利用神经网络方法进行粒子鉴别的研究

本文描述了一种多层感知器的神经网络系统在BESIII粒子鉴别技术中的应用。网络按照子探测器分别进行训练, 输出结果可以作为后续网络的输入或者可以为似然函数方法构建概率密度函数。蒙特卡罗模拟样本的检验结果表明, 利用神经网络方法可以在BESIII上获得较好的粒子鉴别效果。     

Particle Identification using Artificial Neural Networks at BESIII

A multilayered perceptrons neural network technique has been applied in the particle identification at BESIII. The networks are trained in each sub-detector level. The NN output of sub-detectors can be sent to a sequential network or be constructed as PDFs for a likelihood. Good muon-ID, electron-ID and hadron-ID are obtained from the networks by using the simulated Monte Carlo samples.     

Particle identification using artificial neural networks at BESⅢ

A multilayered perceptrons' neural network technique has been applied in the particle identification at BESⅢ. The networks are trained in each sub-detector level. The NN output of sub-detectors can be sent to a sequential network or be constructed as PDFs for a likelihood. Good muon-ID, electron-ID and hadron-ID are obtained from the networks by using the simulated Monte Carlo samples.     

Superpositional Quantum Network Topologies

We introduce superposition-based quantum networks composed of (i) the classical perceptron model of multilayered, feedforward neural networks and (ii) the algebraic model of evolving reticular quantum structures as described in quantum gravity. The main feature of this model is moving from particular neural topologies to a quantum metastructure which embodies many differing topological patterns. Using quantum parallelism, training is possible on superpositions of different network topologies. As a result, not only classical transition functions, but also topology becomes a subject of training. The main feature of our model is that particular neural networks, with different topologies, are quantum states. We consider high-dimensional dissipative quantum structures as candidates for implementation of the model.     

基于FOA-LM算法的超声回波信号参数估计

在超声回波参数估计中,搜索莱文伯格一马夸特(Levenberg-Marquard,LM)算法的最优解会受到迭代初值与参数向量真实解接近程度的影响。针对LM算法对迭代初值敏感的问题,提出了果蝇优化算法(Fruit fly optimization algorithm,FOA)算法和LM算法结合的参数估计方法。该方法充分利用FOA算法善于进行全局搜索和LM算法善于进行局部快速搜索的优点,首先使用FOA算法求出超声回波信号的参数初值,然后利用这组初值进行LM法迭代搜索。仿真结果表明,基于FOA和LM算法相结合的方法,具有收敛速度快,精度高的特点。     

基于邻域粗糙集和并行神经网络的故障诊断

针对目前使用神经网络诊断故障时出现的输入向量选择困难、网络结构复杂、对并发故障诊断效果不好等问题,提出了基于邻域粗糙集和并行神经网络的故障诊断方法。先利用邻域粗糙集对初始征兆进行约简,留下有价值的征兆作为神经网络的输入向量,然后针对每种故障类型设计一个神经网络。用多个训练好的神经网络来并行诊断故障,综合每个神经网络的结果给出最终的诊断结论。用转子实验台的实验数据对这种故障诊断方法进行验证,结果显示该方法能优化神经网络结构,且神经网络具有训练速度快、诊断正确率高的特点。     

Modeling of Thermal Conductivity of Concrete with Vermiculite by Using Artificial Neural Networks Approaches

In this article, the thermal conductivity of concrete with vermiculite is determined and also predicted by using artificial neural networks approaches, namely the radial basis neural network and multi-layer perceptron. In these models, 20 datasets were used. For the training set, 12 datasets (60%) were randomly selected, and the residual datasets (8 datasets, 40%) were selected as the test set. The root mean square error, the mean absolute error, and determination coefficient statistics are used as evaluation criteria of the models, and the experimental results are compared with these models. It is found that the radial basis neural network model is superior to the other models.     

基于深度聚类的目标细粒度分类方法北大核心CSCD

为提高光电系统对弱小目标的识别和分类能力,降低算法对硬件平台和数据的依赖,提出一种无监督分类方法−基于目标深度特征聚类的细粒度分类方法。该方法通过轮廓、颜色、对比度等浅层特征提取提示目标,经超分辨处理后,利用卷积神经网络对目标的深层特征进行编码,进一步采用基于注意机制的主成分分析方法进行降维生成表征矩阵,最后利用聚类的方式实现目标细粒度分类。实验验证了基于不同神经网络的深度聚类方法在不同数据集上的分类性能,其中采用ResNet-34聚类方法在CIFAR-10测试集上细粒度分类性能达92.71%,结果表明,基于深度聚类的目标细粒度方法能够取得与强监督学习方法相当的目标分类效果。此外,还可以根据不同簇数和聚类等级的选择实现不同细粒度的分类效果。     

Temperature measurements of high-temperature semi-transparent infrared material using multi-wavelength pyrometry

Temperature measurements inside semi-transparent materials are important in many fields. This study investigates the measurements of interior temperature distributions in a one-dimensional semi-transparent material using multi-wavelength pyrometry based on the Levenberg–Marquardt method (LMM). The investigated material is semi-transparent Zinc Sulfide (ZnS), an infrared-transmitting optical material operating at long wavelengths. The radiation properties of the one-dimensional semi-transparent ZnS plate, including the effective spectral–directional radiation intensity and the proportion of emitted radiation, are numerically discussed at different wavelengths (8.0–14.0 μm) and temperature distributions (400–800 K) to provide the basic data for the temperature inversion problem. Multi-wavelength pyrometry was combined with the Levenberg–Marquardt method to resolve the temperature distribution along the radiative transfer direction based on the line-of-sight spectral radiation intensities at multiple wavelengths in the optimized spectral range of (11.0–14.0 μm) for the semi-transparent ZnS plate. The analyses of the non-linear inverse problem show that with less than 5.0% noise, the inversion temperature results using the Levenberg–Marquardt method are satisfactory for linear or Gaussian temperature distributions in actual applications. The analysis provides valuable guidelines for applications using multi-wavelength pyrometry for temperature measurements of semi-transparent materials.     

基于CAN总线的高可靠成像控制系统设计

为了实现多套成像设备的智能化控制,设计了基于CAN总线的成像控制系统,并给出了设计中关键技术的解决方法;首先,给出了可靠性高的热备份CAN总线控制系统硬件设计原理;其次,介绍了基于FPGA的CAN总线协议芯片—SJA1000逻辑控制原理与方法;最后结合实际工程项目阐述了控制系统的工作过程;试验结果表明,该设计性能稳定、可靠性高,能够满足多台成像系统的智能化控制;设计理念和方法具有通用性,系统的可扩展性强。     

Neural estimator to determine alpha parameter in terms of quantum-well number

The value of the α (Alpha) parameter, which is also called Linewidth enhancement factor, is particularly important in optical communication systems. This paper presents a new approach based on artificial neural networks (ANNs) to determine the α parameter for different number of quantum-wells (QWs). ANNs are trained in different structures with the use of five learning algorithms to obtain better performance and faster error convergence. The Levenberg–Marquardt (LM) algorithm, which has a quadratic speed of convergence, gives the best result among other learning algorithms used in the analysis. Both the training and the test results are in very good agreement with the experimental results reported elsewhere.     

Noise reduction for lidar returns using self-adaptive wavelet neural network

Lidar has been widely applied in many fields, such as meteorology and environment. However, because lidar returns are very weak, the influence of noise on useful signal is very serious. To obtain useful lidar return signals from raw data, a self-adaptive method combining wavelet analysis and a neural network that suppresses noise is proposed, in which the orthogonal Daubechies wavelet family serves as node functions in the hidden layer of the neural network, a search algorithm is selected to optimize the parameters and thresholds, and the Levenberg–Marquardt algorithm is adopted in the neural network gradient algorithm. Some comparative experiments were carried out to verify the feasibility of the noise reduction method and the results showed that the signal-to-noise ratio (SNR) of the common wavelet threshold denoising method is about 10, while that of the self-adaptive wavelet neural network denoising method is more than 20. From the experimental results, it can be seen that the wavelet neural network denoising method has less distortion and a higher SNR value than other methods, giving it superior performance.