深孔控制爆破在渝长公路路堑开挖中的应用

介绍了渝长公路K31+4 2 5 - 6 80段路堑开挖深孔爆破的工程实例 ,论述了爆破方案的确定、爆破参数设计及主要的安全措施等。针对复杂的爆破环境 ,采用预裂爆破、加强堵塞和严格控制最大一段起爆药量等技术措施 ,保证了爆破安全 ,取得了预想的爆破效果。     

腊山残留巨石柱超深孔控制爆破

在城市内破损山体治理中,通过采用深孔控制爆破、毫秒延时起爆网路及合理的爆破参数,防止了爆破飞石、降低了爆破振动,保证了爆破效果。本工程成功地对残留巨石柱进行了一次性爆破,可供类似工程借鉴。     

Detection and Classification of Diabetic Retinopathy Using DCNN and BSN Models

Diabetes is associated with many complications that could lead to death. Diabetic retinopathy, a complication of diabetes, is difficult to diagnose and may lead to vision loss. Visual identification of micro features in fundus images for the diagnosis of DR is a complex and challenging task for clinicians. Because clinical testing involves complex procedures and is time-consuming, an automated system would help ophthalmologists to detect DR and administer treatment in a timely manner so that blindness can be avoided. Previous research works have focused on image processing algorithms, or neural networks, or signal processing techniques alone to detect diabetic retinopathy. Therefore, we aimed to develop a novel integrated approach to increase the accuracy of detection. This approach utilized both convolutional neural networks and signal processing techniques. In this proposed method, the biological electro retinogram (ERG) sensor network (BSN) and deep convolution neural network (DCNN) were developed to detect and classify DR. In the BSN system, electrodes were used to record ERG signal, which was pre-processed to be noise-free. Processing was performed in the frequency domain by the application of fast Fourier transform (FFT) and mel frequency cepstral coefficients (MFCCs) were extracted. Artificial neural network (ANN) classifier was used to classify the signals of eyes with DR and normal eye. Additionally, fundus images were captured using a fundus camera, and these were used as the input for DCNN-based analysis. The DCNN consisted of many layers to facilitate the extraction of features and classification of fundus images into normal images, non-proliferative DR (NPDR) or early-stage DR images, and proliferative DR (PDR) or advanced-stage DR images. Furthermore, it classified NPDR according to microaneurysms, hemorrhages, cotton wool spots, and exudates, and the presence of new blood vessels indicated PDR. The accuracy, sensitivity, and specificity of the ANN classifier were found to be 94%, 95%, and 93%, respectively. Both the accuracy rate and sensitivity rate of the DCNN classifier was 96.5% for the images acquired from various hospitals as well as databases. A comparison between the accuracy rates of BSN and DCNN approaches showed that DCNN with fundus images decreased the error rate to 4%.     

Understanding the electrochemical behavior of Sn(II) in choline chloride-ethylene glycol deep eutectic solvent for tin powders preparation

Tin is one of the vital substances in the material and chemical industry. Deep eutectic solvent (DES) possesses excellent physicochemical properties for metal preparation. In this study, the electrochemical behavior of Sn(II) and metallic tin preparation in choline chloride-ethylene glycol (ChCl-EG) DES were investigated. Linear sweep voltammetry is carried out to study the effects of substrate material (Ni, Cu, Pt, C, Fe, Al, W, glassy carbon (GC) and Ti), temperature (313–353 K) and SnCl₂ concentration (10–100 mM) on the electrochemical behavior of Sn(II)/Sn. Results demonstrate that metallic tin is easier to be prepared on the Ni electrode, and increasing temperature and SnCl₂ concentration both facilitate the reduction of Sn(II). Electrospray ionization mass spectrometry (ESI-MS) shows that the main form of Sn(II) is [SnCl₃]^-. Chronoamperometric tests indicate that with the increase of applied potential, the nucleation of Sn(II) on the nickel electrode becomes close to the instantaneous nucleation. Simultaneously, the nucleation rate and the number of active sites on the electrode surface increase significantly. Pure tin with a dendritic structure could be prepared with higher current efficiency of 96.4% and lower energy consumption of 320.5 kW·h·t^?1 without any additives. All these studies provide a fundamental basis for the preparation of tin powders in ChCl-EG DES.     

DBN在蛋白质编码区识别问题中的应用研究

针对真核生物DNA序列中蛋白质编码区的识别问题,提出基于深度置信网络（Deep Belief Network,DBN）的组合模型。通过信号处理技术对真核生物的DNA序列进行数值转换,并结合统计学知识提取转换后DNA序列的数值特征;利用随机森林对所提取的特征变量降维;用深度置信网络模型对DNA序列分类判别;根据短时傅里叶变换（Short Time Fourier Transform,STFT）技术对外显子区准确定位。在三个标准测试集上比较组合模型与传统[Logistic]回归模型、贝叶斯判别模型的判别效果,结果显示,深度置信网络组合模型的准确率和特异度等指标都明显优于[Logistic]回归模型和贝叶斯判别模型。     

基于CNN的恶意Web请求检测技术

目前,基于卷积神经网络的Web恶意请求检测技术领域内只有针对URL部分进行恶意检测的研究,并且各研究对原始数据的数字化表示方法不同,这会造成检测效率和检测准确率较低。为提高卷积神经网络在Web恶意请求检测领域的性能,在现有工作的基础上将其他多个HTTP请求参数与URL合并,将数据集HTTP data set CSIC 2010和DEV_ACCESS作为原始数据,设计对比实验。首先采用6种数据数字向量化方法对字符串格式的原始输入进行处理;然后将其分别输入所设计的卷积神经网络,训练后可得到6个不同的模型,同时使用相同的训练数据集对经典算法HMM,SVM和RNN进行训练,得到对照组模型;最后在同一验证集上对9个模型进行评估。实验结果表明,采用多参数的Web恶意请求检测方法将词汇表映射与卷积神经网络内部嵌入层相结合对原始数据进行表示,可使卷积神经网络取得99.87%的准确率和98.92%的F1值。相比其他8个模型,所提方法在准确率上提升了0.4~7.7个百分点,在F1值上提升了0.3~13个百分点。实验充分说明,基于卷积神经网络的多参数Web恶意请求检测技术具有明显的优势,且使用词汇表映射和网络内部嵌入层对原始数据进行处理能使该模型取得最佳的检测效果。     

基于深度神经网络的“弱监督”密集场景人群计数算法

目前,在密集场景人群计数任务中,标注真实密度图的方法是对行人头部的中心位置进行标注,并利用高斯卷积生成真实的密度分布图作为监督信息。但是,对于密集场景而言,这样的标注方式是费时、费力的,并且密集场景图片中有诸多“非受控”因素,如低分辨率、背景噪声、目标遮挡和尺度变化等。针对这一问题,提出了一种新的标注方法,即只需要知道图片中包含多少个物体,以图片中行人的数量作为监督信息。与传统的真实密度图相比,所提出的标记方法中以真实目标的数值为“弱监督”信息。实验结果表明,对于人群回归任务,利用弱监督信息对神经网络进行训练得到的模型能够较为准确地回归出图片中所包含目标的数量,从而证明了该方法的有效性。     

基于深度学习的隐式篇章关系识别综述

隐式篇章关系识别是自然语言处理中一项富有挑战性的任务,旨在判断缺少连接词的两个论元(子句或者句子)之间的语义关系(例如转折)。近年来,随着深度学习在自然语言处理领域的广泛应用,各种基于深度学习的隐式篇章关系识别方法取得了不错的效果,其性能全面超越了早期基于人工特征的方法。文中分三大类对最近的隐式篇章关系识别方法进行讨论:基于论元编码的方法、基于论元交互的方法和引入显式篇章数据的半监督方法。在PDTB数据集上的实验结果显示:1)通过显式地建模论元中词或文本片段之间的语义关系,基于论元交互的方法的性能明显好于基于论元编码的方法;2)引入显式篇章数据的半监督方法能有效地缓解数据稀疏问题,从而进一步提升识别的性能。最后,分析了当前面临的主要问题,并指出了未来可能的研究方向。     

基于深度学习的多标签生成研究进展

大数据时代,数据呈现维度高、数据量大和增长快等特点。如何有效利用其中蕴含的有价值信息,以实现数据的智能化处理,已成为当前理论和应用的研究热点。针对现实普遍存在的多义性对象,数据多标签被提出并被广泛应用于数据智能化组织。近年来,深度学习在数据特征提取方面呈现出高速、高精度等优异性,使基于深度学习的多标签生成得到广泛关注。文中分五大类别总结了最新研究成果,并进一步从数据、关系类型、应用场景、适应性及实验性能方面对其进行对比和分析,最后探讨了多标签生成面临的挑战和未来的研究方向。     

基于深度学习网络模型的端到端航迹关联

为提高雷达数据处理中航迹关联的智能性,充分利用目标的特征信息,并简化系统处理流程,提出了一种基于深度学习网络模型的端到端航迹关联算法。首先分析了基于神经网络的航迹关联存在样本细节少、处理流程繁杂的问题,然后提出了端到端的深度学习模型。该模型根据航迹关联数据的处理特征,改进了卷积神经网络结构用于特征提取,充分利用了长短期记忆网络对历史信息和将来信息的处理能力,并分析了前后航迹的关联性。在对原始数据进行卡尔曼滤波后,将全部航迹信息特征作为输入,并由基于卷积神经网络特征提取的长短期记忆深度神经网络模型直接输出航迹关联结果。仿真结果表明,提出的模型可以充分学习推演目标的多个特征信息,具有较高的航迹关联准确率,对航迹关联的智能化分析具有一定的参考价值。