Failure load prediction and optimisation for adhesively bonded joints enabled by deep learning and fruit fly optimisation

Adhesively bonded joints have been extensively employed in the aeronautical and automotive industries to join thin-layer materials for developing lightweight components. To strengthen the structural integrity of joints, it is critical to estimate and improve joint failure loads effectually. To accomplish the aforementioned purpose, this paper presents a novel deep neural network (DNN) model-enabled approach, and a single lap joint (SLJ) design is used to support research development and validation. The approach is innovative in the following aspects: (i) the DNN model is reinforced with a transfer learning (TL) mechanism to realise an adaptive prediction on a new SLJ design, and the requirement to re-create new training samples and re-train the DNN model from scratch for the design can be alleviated; (ii) a fruit fly optimisation (FFO) algorithm featured with the parallel computing capability is incorporated into the approach to efficiently optimise joint parameters based on joint failure load predictions. Case studies were developed to validate the effectiveness of the approach. Experimental results demonstrate that, with this approach, the number of datasets and the computational time required to re-train the DNN model for a new SLJ design were significantly reduced by 92.00% and 99.57% respectively, and the joint failure load was substantially increased by 9.96%.     

A Strategy of Signal Detection for Performance Improvement in Clipping Based OFDM System

In this paper, the supervised Deep Neural Network (DNN) based signal detection is analyzed for combating with nonlinear distortions efficiently and improving error performances in clipping based Orthogonal Frequency Division Multiplexing (OFDM) ssystem. One of the main disadvantages for the OFDM is the high Peak to Average Power Ratio (PAPR). The clipping is a simple method for the PAPR reduction. However, an effect of the clipping is nonlinear distortion, and estimations for transmitting symbols are difficult despite a Maximum Likelihood (ML) detection at the receiver. The DNN based online signal detection uses the offline learning model where all weights and biases at fullyconnected layers are set to overcome nonlinear distortions by using training data sets. Thus, this paper introduces the required processes for the online signal detection and offline learning, and compares error performances with the ML detection in the clipping-based OFDM systems. In simulation results, the DNN based signal detection has better error performance than the conventional ML detection in multi-path fading wireless channel. The performance improvement is large as the complexity of system is increased such as huge Multiple Input Multiple Output (MIMO) system and high clipping rate.     

基于DNN／NHibernate的Web—MIS设计与实现

引入了基于．NET环境的2种具有不同作用的框架：DNN与NHibemate，分析了DNN和NHibemate的体系结构，构建了基于DNN／NHibemate的Web-MIS的框架结构，给出了系统各层的具体实现方法．某Web-MIS的开发实践证明，利用DNN／NHibemate框架开发Web-MIS可提高系统的开发效率和可靠性．     

基于困惑度数据挑选的半监督声学建模*

针对资源稀少情况下小语种的声学建模问题,提出根据解码后文本的困惑度挑选无监督数据并重新训练声学模型的策略.使用少量精标数据训练得到一个初始种子模型后,解码大量无监督数据,计算解码后的文本与精标数据文本的困惑度,从中挑选与精标数据相近的数据,再将这些数据与原有精标数据共同用于声学模型训练.为了提高解码的无监督数据的正确性,在基于深层神经网络的模型参数训练中,当最后一次模型参数更新时只使用精标数据修正网络参数.在NIST 2015年关键词识别比赛中Swahili语的VLLP识别任务上,相比其它方法,文中方法的识别率有一定提升.     

基于深度神经网络和门控循环单元的动态图表示学习方法

学习图中节点的潜在向量表示是一项重要且普遍存在的任务,旨在捕捉图中节点的各种属性。大量工作证明静态图表示已经能够学习到节点的部分信息,然而,真实世界的图是随着时间的推移而演变的。为了解决多数动态网络算法不能有效保留节点邻域结构和时态信息的问题,提出了基于深度神经网络（DNN）和门控循环单元（GRU）的动态网络表示学习方法DynAEGRU。该方法以自编码器作为框架,其中的编码器首先用DNN聚集邻域信息以得到低维特征向量,然后使用GRU网络提取节点时态信息,最后用解码器重构邻接矩阵并将其与真实图对比来构建损失。通过与几种静态图和动态图表示学习算法在3个数据集上进行实验分析,结果表明DynAEGRU具有较好的性能增益。     

Non-Contact Physiological Measurement System for Wearing Masks During the Epidemic

Physiological signals indicate a person’s physical and mental state at any given time. Accordingly, many studies extract physiological signals from the human body with non-contact methods, and most of them require facial feature points. However, under COVID-19, wearing a mask has become a must in many places, so how non-contact physiological information measurements can still be performed correctly even when a mask covers the facial information has become a focus of research. In this study, RGB and thermal infrared cameras were used to execute non-contact physiological information measurement systems for heart rate, blood pressure, respiratory rate, and forehead temperature for people wearing masks due to the pandemic. Using the green (G) minus red (R) signal in the RGB image, the region of interest (ROI) is established in the forehead and nose bridge regions. The photoplethysmography (PPG) waveforms of the two regions are obtained after the acquired PPG signal is subjected to the optical flow method, baseline drift calibration, normalization, and bandpass filtering. The relevant parameters in Deep Neural Networks (DNN) for the regression model can correctly predict the heartbeat and blood pressure. In addition, the temperature change in the ROI of the mask after thermal image processing and filtering can be used to correctly determine the number of breaths. Meanwhile, the thermal image can be used to read the temperature average of the ROI of the forehead, and the forehead temperature can be obtained smoothly. The experimental results show that the above-mentioned physiological signals of a subject can be obtained in 6-s images with the error for both heart rate and blood pressure within 2%~3% and the error of forehead temperature within ±0.5°C.     

基于深度神经网络的水声FBMC通信信号检测方法

针对传统水声滤波器组多载波（Filter Bank Multi-Carrier,FBMC）通信接收端需经过信道估计和均衡才可恢复出发送符号,系统复杂度高且信道估计精度不佳等问题。文章将深度神经网络融入到水声多载波通信当中,提出一种基于深度神经网络的水声FBMC信号检测方法。在训练阶段通过大量的数据迭代、调试超参数和优化算法来改善深度神经网络参数,使其具有预期的估计效果。利用训练完成的深度神经网络模型取代传统FBMC通信系统接收端的信道估计、均衡等模块,自适应地学习水声信道状态信息,同时避免了固有的虚部干扰影响。在测试阶段直接将频域序列作为网络的输入来预测发送的二进制序列,仿真结果表明所提出的基于深度神经网络的FBMC信号检测方法相比传统信道估计算法有更好的误码率性能。     

深度神经网络在螺旋桨叶片数识别中的应用

从调制(Demodulation on Noise, DEMON)谱谐波簇中提取的结构特征可以建立用于螺旋桨叶片数识别的模板。使用模板匹配算法进行螺旋桨叶片数识别时,存在依赖模板库和置信度准则、算法约束条件多、无法发现缺失模板等问题。本文提出了一种将深度神经网络(Deep Neural Network, DNN)应用于螺旋桨叶片数识别的方法,该方法仅在训练深度神经网络时使用模板库,克服了识别过程中对模板库和置信度准则的依赖。此外,通过提取识别错误项,可以找到缺失模板,实现了对模板库数据的补充。使用该算法对大量实测数据进行检测,发现深度神经网络具有更高的识别正确率,而且识别过程更加简单可靠。     

Bi-directional LSTM based channel estimation in 5G massive MIMO OFDM systems over TDL-C model with Rayleigh fading distribution

In this work, a deep learning (DL)-based massive multiple-input multiple-output (mMIMO) orthogonal frequency division multiplexing (OFDM) system is investigated over the tapped delay line type C (TDL-C) model with a Rayleigh fading distribution at frequencies ranging from 0.5 to 100 GHz. The proposed bi-directional long short-term memory (Bi-LSTM) channel state information (CSI) estimator uses online learning during training and offline learning during the practical implementation phase. The design of the estimator takes into account situations in which prior knowledge of channel statistics is limited and targets excellent performance, even with limited pilot symbols (PS). Three separate loss functions (mean square logarithmic error [MSLE], Huber, and Kullback–Leibler Distance [KLD]) are assessed in three classification layers. The symbol error rate (SER) and outage probability performance of the proposed estimator are evaluated using a number of optimization techniques, such as stochastic gradient descent (SGD), momentum, and the adaptive gradient (AdaGrad) algorithm. The Bi-LSTM-based CSI estimator is trained considering a specific number of PS. It can be readily seen that by incorporating a cyclic prefix (CP), the system becomes more resilient to channel impairments, resulting in a lower SER. Simulations show that the SGD optimization approach and Huber loss function-trained Bi-LSTM-based CSI estimator have the lowest SER and very high estimation accuracy. By using deep neural networks (DNNs), the Bi-LSTM method for CSI estimation achieves a superior channel capacity (in bps/Hz) at 10 dB than long short-term memory (LSTM) and other conventional CSI estimators, such as minimum mean square error (MMSE) and least squares (LS). The simulation results validate the analytical results in the study.     

基于资源管理的信息融合系统工程的研究

基于资源管理的信息融合系统开发是信息融合系统工程相关技术发展的新方向.资源管理问题贯穿信息融合整个过程,在对信息融合中的资源管理问题分析的基础上,提出并分析了基于资源管理的功能模型,并在此基础上进一步分析了将资源管理用于信息融合的DNN结构,该结构形式简单易于操作,为后续操作提供了便利.在此基础上,对基于资源管理的信息融合系统工程方法进行了研究,给出了基于资源管理的信息融合系统工程过程,有助于提高信息融合工程的经济性,并对自适应信息融合的实现、多级数据融合的协调具有重要意义.