UWB LOS/NLOS identification in multiple indoor environments using deep learning methods

Indoor location-aware service is booming in daily life and business activities, making the demand for precise indoor positioning systems thrive. The identification between line-of-sight (LOS) and non-line-of-sight (NLOS) is critical for wireless indoor time-of-arrival-based localization methods. Ultra-Wide-Band (UWB) is considered low cost among the many wireless positioning systems. It can resolve multi-path and have high penetration ability. This contribution addresses UWB NLOS/LOS identification problem in multiple environments. We propose a LOS/NLOS identification method using Convolutional Neural Network parallel with Gate Recurrent Unit, named Indoor NLOS/LOS identification Neural Network. The Convolutional Neural Network extracts spatial features of UWB channel impulse response data. While the Gate Recurrent Unit is an effective approach for designing deep recurrent neural networks which can extract temporal features. By integrating squeeze-and-extraction blocks into these architectures we can assign weights on channel-wise features. We simulated UWB channel impulse response signals in residential, office, and industrial scenarios based on the IEEE 802.15.4a channel model report. The presented network was tested in simulation scenarios and an open-source real-time measured dataset. Our method can solve NLOS identification problems for multiple indoor environments. Thus more versatile compare with networks only working in one scenario. Popular machine learning methods and deep learning methods are compared against our method. The test results show that the proposed network outperforms benchmark methods in simulation datasets and real-time measured datasets.     

Fast neutron spectra determination by threshold activation detectors using neural networks

Neural network method was used for fast neutron spectra unfolding in spectrometry by threshold activation detectors. The input layer of the neural networks consisted of 11 neurons for the specific activities of neutron-induced nuclear reaction products, while the output layers were fast neutron spectra which had been subdivided into 6, 8, 10, 12, 15 and 20 energy bins. Neural network training was performed by 437 fast neutron spectra and corresponding threshold activation detector readings. The trained neural network have been applied for unfolding 50 spectra, which were not in training sets and the results were compared with real spectra and unfolded spectra by SANDII. The best results belong to 10 energy bin spectra. The neural network was also trained by detector readings with 5% uncertainty and the response of the trained neural network to detector readings with 5%, 10%, 15%, 20%, 25% and 50% uncertainty was compared with real spectra. Neural network algorithm, in comparison with other unfolding methods, is very fast and needless to detector response matrix and any prior information about spectra and also the outputs have low sensitivity to uncertainty in the activity measurements. The results show that the neural network algorithm is useful when a fast response is required with reasonable accuracy.     

Quantum Competitive Neural Network

Quantum Neural Network (QNN) is a fledging science built upon the combination of classical neural network and quantum computing. After analyzing of traditional competitive neural network, this paper firstly presents a Quantum Competitive Neural Network (QCNN) that can recognize patterns and classify patterns via quantum competition. Contrasting to the conventional competitive neural network, the storage capacity or memory capacity of the QCNN is exponentially increased by a factor of 2 ⁿ , where n is the number of qubit. The QCNN has no weights, does not need to learn and update weights, which accelerates the learning process of the network. Besides, the case analysis validates the feasibility and validity of the QCNN in this paper.     

Image feature extraction by dynamic neural filtering and phase-only joint transform correlation

Neural network-based image processing algorithms present numerous advantages due to their supervised adjustable properties. Among various neural network architectures, dynamic neural networks, Hopfield and Cellular networks, have been found inherently suitable for filtering applications. Combining supervised and filtering features of dynamic neural networks, this paper presents dynamic neural filtering technique based on Hopfield neural network architecture. The filtering technique has also been implemented by using phase-only joint transform correlation (POJTC) for optical image processing applications. Filtering structure is basically similar to the Hopfield neural network structure except for the adjustable filter mask and 2D convolution operation instead of weight matrix operations. The dynamic neural filtering architecture has learnable properties by back-propagation learning algorithm. POJTC presents significant advantages to achieve the operation of summing the cross-correlation of bipolar data by phase-encoding bipolar data in parallel. The image feature extraction performance of the proposed optical system is reported for various image processing applications using a simulation program.     

基于RBF网络的光纤荧光海水叶绿素a含量在线监测系统的研究

介绍了能够实现对海水中叶绿素a含量在线监测的光纤荧光系统。系统将荧光技术、光纤技术和基于RBF(Radial Basis Function Neural Network)神经网络的标定方法相结合，建立了用于海水叶绿素a含量在线监测的最佳RBF网络结构，系统具有结构简单，探头无源及高灵敏度等特点。     

Design of Nonlinear Autoregressive Exogenous Model Based Intelligence Computing for Efficient State Estimation of Underwater Passive Target

In this study, an intelligent computing paradigm built on a nonlinear autoregressive exogenous (NARX) feedback neural network model with the strength of deep learning is presented for accurate state estimation of an underwater passive target. In underwater scenarios, real-time motion parameters of passive objects are usually extracted with nonlinear filtering techniques. In filtering algorithms, nonlinear passive measurements are associated with linear kinetics of the target, governing by state space methodology. To improve tracking accuracy, effective feature estimation and minimizing position error of dynamic passive objects, the strength of NARX based supervised learning is exploited. Dynamic artificial neural networks, which contain tapped delay lines, are suitable for predicting the future state of the underwater passive object. Neural networks-based intelligence computing is effectively applied for estimating the real-time actual state of a passive moving object, which follows a semi-curved path. Performance analysis of NARX based neural networks is evaluated for six different scenarios of standard deviation of white Gaussian measurement noise by following bearings only tracking phenomena. Root mean square error between estimated and real position of the passive target in rectangular coordinates is computed for evaluating the worth of the proposed NARX feedback neural network scheme. The Monte Carlo simulations are conducted and the results certify the capability of the intelligence computing over conventional nonlinear filtering algorithms such as spherical radial cubature Kalman filter and unscented Kalman filter for given state estimation model.     

正则化参数求解方法研究

针对BP神经网络算法训练过程中出现的过拟合问题,提出了利用一阶原点矩,二阶原点矩,方差和极大似然估计概念的推广来计算L2正则化中正则化参数 值的方法。该方法通过对算法数据集[X,Y]中的X矩阵进行运算得到的四个 值,BP神经网络算法训练时通常采用的是贝叶斯正则化方法,贝叶斯正则化方法存在着对先验分布和数据分布依赖等问题,而利用上述概念的推广计算的参数代入L2正则化的方法简便没有应用条件限制。在BP神经网络手写数字识别的实验中,将该方法与贝叶斯正则化方法应用到实验中后的算法识别结果进行比较,正确率提高了1.14-1.50个百分点。因而计算得到的 值应用到L2正则化方法与贝叶斯正则化方法相比更能使得BP神经网络算法的泛化能力强,证明了该算法的有效性。     

Simultaneous Determination of Multicomponents in Air Toxic Organic Compounds Using Artificial Neural Networks in Ftir Spectroscopy

The application of Artificial Neural Networks (ANNs) for nonlinear multivariate calibration using simulated FTIR data was demonstrated in this paper. Neural networks consisting of three layers of nodes were trained by using the back-propagation learning rule. Since parameters affect the performance of the network greatly, simulated data were used to train the network in order to get a satisfactory combination of all parameters. The mixtures of four air toxic organic compounds whose FTIR spectra are overlapped were chosen to evaluate the calibration and prediction ability of the network. The relative standard error (RSD%), the percent standard error of prediction samples (%SEP) and the percent standard error of calibration samples (%SEC) are used for evaluating the ability of the neural network.     

人工神经网络在超声无损检测中的应用

人工神经网络对于超声无损检测的发展具有重要的意义，本文对国内外人工神经网络在超声无损检测中应用研究做了较为详细的介绍和分析，指出人工神经网络是实现超声无损检测测定量化的有效途径。     

基于改进BP神经网络的航空液压油软测量

BP 神经网络越来越多地被应用于软测量建模中,与传统方法相比,BP神经网络进行信息处理可以减少数据的分析和建模工作,但也存在易于陷入局部最小值和初始权值随机选取的缺陷问题。为了解决传统BP神经网络存在的缺陷,论文中在数据预处理过程中引入主成分分析法（PCA）,在BP网络输入权值时引入遗传算法（GA）,最终达到弥补BP神经网络缺陷的目的;详细介绍了改进算法的流程与步骤,将改进的BP神经网络应用于航空液压油的软测量,先是对航空液压油软测量参数进行分析,包括辅助变量的选择和数据预处理,然后进行基于改进型BP神经网络的建模与仿真实验。实验结果表明,基于改进BP神经网络的航空液压油软测量效果优于传统神经网络,具有更强的泛化能力,因此可进行更广泛的应用。     

人工神经网络法鉴别红外光谱

本文将反向传播人工神经网络（ＢＰ－ＡＮＮ）用于ＦＴＩＲ,鉴别未知化合物。结果表明,当训练集样本不含噪声时,纯光谱的预测结果很好。而当训练集样本有少量噪声干扰时,预测结果随预测集样本的不同,而得到不同的改善。     

Soliton,breather, and rogue wave solutions for solving the nonlinear Schrödinger equation using a deep learning method with physical constraints

The nonlinear Schro¨dinger equation is a classical integrable equation which contains plenty of significant properties and occurs in many physical areas.However,due to the difficulty of solving this equation,in particular in high dimensions,lots of methods are proposed to effectively obtain different kinds of solutions,such as neural networks among others.Recently,a method where some underlying physical laws are embeded into a conventional neural network is proposed to uncover the equation’s dynamical behaviors from spatiotemporal data directly.Compared with traditional neural networks,this method can obtain remarkably accurate solution with extraordinarily less data.Meanwhile,this method also provides a better physical explanation and generalization.In this paper,based on the above method,we present an improved deep learning method to recover the soliton solutions,breather solution,and rogue wave solutions of the nonlinear Schro¨dinger equation.In particular,the dynamical behaviors and error analysis about the one-order and two-order rogue waves of nonlinear integrable equations are revealed by the deep neural network with physical constraints for the first time.Moreover,the effects of different numbers of initial points sampled,collocation points sampled,network layers,neurons per hidden layer on the one-order rogue wave dynamics of this equation have been considered with the help of the control variable way under the same initial and boundary conditions.Numerical experiments show that the dynamical behaviors of soliton solutions,breather solution,and rogue wave solutions of the integrable nonlinear Schro¨dinger equation can be well reconstructed by utilizing this physically-constrained deep learning method.     

Neural Networks for Estimating Speculative Attacks Models

Currency crises have been analyzed and modeled over the last few decades. These currency crises develop mainly due to a balance of payments crisis, and in many cases, these crises lead to speculative attacks against the price of the currency. Despite the popularity of these models, they are currently shown as models with low estimation precision. In the present study, estimates are made with first- and second-generation speculative attack models using neural network methods. The results conclude that the Quantum-Inspired Neural Network and Deep Neural Decision Trees methodologies are shown to be the most accurate, with results around 90% accuracy. These results exceed the estimates made with Ordinary Least Squares, the usual estimation method for speculative attack models. In addition, the time required for the estimation is less for neural network methods than for Ordinary Least Squares. These results can be of great importance for public and financial institutions when anticipating speculative pressures on currencies that are in price crisis in the markets.     

非线性系统混沌运动的神经网络控制

设计前馈反传神经网络控制非线性系统混沌运动的新方法.根据扰动参数模型输入输出数据,按照非线性学习算法训练网络产生系统稳定所需的小扰动控制信号,去镇定混沌运动,使嵌入在混沌吸引子中的不稳定周期轨道回到稳定不动点上.Hnon映射数值仿真结果表明,这种方法控制非线性混沌系统响应速度快、控制精度高 关键词： 混沌控制 神经网络 吸引子 非线性     

Adaptive Fixed-Time Neural Network Tracking Control of Nonlinear Interconnected Systems

In this article, a novel adaptive fixed-time neural network tracking control scheme for nonlinear interconnected systems is proposed. An adaptive backstepping technique is used to address unknown system uncertainties in the fixed-time settings. Neural networks are used to identify the unknown uncertainties. The study shows that, under the proposed control scheme, each state in the system can converge into small regions near zero with fixed-time convergence time via Lyapunov stability analysis. Finally, the simulation example is presented to demonstrate the effectiveness of the proposed approach. A step-by-step procedure for engineers in industry process applications is proposed.     

Application of functional link artificial neural network for mitigating nonlinear effects in coherent optical OFDM

Artificial Neural Networks are developed as an important technique for equalization and have been widely used to mitigate the nonlinear effects in coherent optical systems. For the compensation of nonlinearities in coherent optical orthogonal frequency division multiplexing technique, the most popular artificial neural network model is a multilayer perceptron (MLP), as it is able to perform complex mapping between input and output spaces with significant success. However due to the complexity of multilayer perceptron nonlinear equalizer (MLP-NLE) model training of neural network is difficult. To overcome computational complexity issues of MLP-NLE, a single neuron based functional link artificial neural network nonlinear equalizer (FLANN-NLE) has been developed in this paper. Better performance of an equalizer is attributed to the usage of aPSO-BP algorithm for training the FLANN-NLE. The proposed FLANN-NLE surpasses the existing works both in terms of Q-Factor and computational complexity. For a fiber length of 1000 km and at launch power of ?6 dBm, the improvement in Q-Factor is approximately equal to 3.3 and 1 dB in contrast to the previously reported values of approximately 3 and 0.7 dB at bit rate of 40 and 80 Gbps respectively.     

Topology group concept for truss topology optimization with frequency constraints

A practical methodology based on a topology group concept is presented for finding optimal topologies of trusses. The trusses are subjected to natural frequency, stress, displacement and Euler buckling constraints. Multiple loading conditions are considered, and a constant nodal mass is assumed for each existing node. The nodal cost as well as the member cost is incorporated in the cost function. Starting with a ground structure, a sequence of substructures with different node distribution, called topology group, is generated by using the binary number combinatorial algorithm. Before optimizing a certain topology, its meaningfulness should be examined. If a topology is meaningless, it is then excluded; otherwise, it is optimized as a sectional area optimization problem. In order to avoid a singular solution, the dimension of the structure for a given topology is kept unchanged in the optimization process by giving the member to be removed a tiny sectional area. A parabolic interpolation method is used to solve a non-linear constrained problem, which forms the part of the algorithm. The efficiency of the proposed method is demonstrated by two typical examples of truss.     

Quantum M-P Neural Network

Quantum Neural Network (QNN) is a young and outlying science built upon the combination of classical neural network and quantum computing. Making use of quantum linear superposition, this paper presents a quantum M-P neural network based on the analysis of the conventional M-P neural network. Moreover, the working principle of this proposed network and its corresponding weight updating algorithm are expatiated in the two cases of input state being in the orthogonal and non-orthogonal basic set, respectively. In addition, this paper not only validates that this quantum M-P network can realize some network functions, such as “XOR”, but also verifies the feasibility and validity of its weight learning algorithm by some simple examples.     

基于SCFNN之PAM非线性信道均衡器成效研究

自组织型模糊类神经网络(SCFNN)可依据一定的法则自我构建神经网络的组织结构,从而适用于当前控制对象;多层神经元是传统的类神经网络,广泛应用于各个领域;倒传递学习法与最陡坡降法相结合,可使以上两种类神经网络进行有效的融合;目前,信道均衡器上的系统架构种类非常多,各种类神经网络应用于信道均衡器也颇为普遍;在研究SCFNN的基础上,将其应用于通道均衡器确实可行,效果良好;比较了SCFNN与MLP在通道均衡器的成效;仿真表明,在相同通道环境下,SCFNN的训练收敛速度、位错误率与系统敏感度优于MLP,完成结构学习后SCFNN的结构也颇为精简。     

Micropositioning using shape memory alloy actuators

Shape memory alloys (SMA) can be used to generate motion or force in electromechanical devices and micro-machines, although their accuracy is severely limited by their highly nonlinear and hysteretical stimulus-response characteristics. In this work we present some results regarding a nonlinear control method suitable for SMA-based positioning applications. In particular, we show how the hysteresis effects can be compensated using an inverse hysteresis model generated by a neural network, trained using experimental data. The control strategy, experimented on a laboratory SMA actuator, uses the inverse model inserted in a proportional-integral with antiwindup control loop. It is found that neural networks successfully improve the closed-loop response, leading to position accuracies close to a micrometer.