Predicting Compressive Strength of Recycled Concrete for Construction 3D Printing Based on Statistical Analysis of Various Neural Networks

Construction 3D printing is changing construction industry, but for its immaturity, there are still many problems to be solved. One of the major problems is to study materials for construction 3D printing. Because printed buildings are very different from traditional buildings, there are special requirements for printing materials. Based on environmental and cost considerations, the recycled concrete as printing material is a perfect choice. In order to study and develop the construction 3D printing materials, it is necessary to predict the properties of them. As one of the most effective artificial intelligence algorithms, artificial neural network can deal with multi-parameter and nonlinear problems, and it can provide useful reference to predict the performance of recycled concrete for 3D printing. However, since there are many types and parameters for neural network, it is difficult to select the optimal neural network with excellent prediction performance. In this paper, by comparing different types of neural networks and statistically analyzing the distribution of the root-mean-square error (RMSE) and the coefficient of determination (R2) of these neural networks, we can determine the best performance among four neural networks and finally select the suitable one to predict the performance of 3D printing concrete.

基于改进LSTM的数码雷管模组印刷质量预测

由于数码雷管模组印刷过程中生产工艺复杂、强时序性等特点,其质量的精准预测已成为提高产品质量管理水平的关键。基于此提出一种改进长短期记忆(long short-term memory,LSTM)网络的数码雷管模组印刷质量预测模型。首先根据数码雷管模组印刷过程提炼机器运行参数、环境参数与检测参数作为印刷产品质量的原始特征,并对关键检测参数进行时序特征重构以增强特征表达能力;其次基于改进的LSTM网络建立数码雷管模组印刷特征提取框架,采用卷积神经网络提取空间特征避免LSTM挖掘高维印刷特征时隐含关系的不足,通过全局注意力机制自适应学习不同时刻印刷特征对印刷产品质量的贡献度,为LSTM提取的深层时序特征分配不同权值;最后以深层特征作为输入,通过全连接网络实现数码雷管模组印刷产品的质量预测。实验结果表明,相较于BP神经网络、门控循环单元网络、LSTM等预测方法,改进的LSTM网络有效提高了数码雷管模组印刷产品质量的预测精度。

E-print网络学术资源初探

E- print是随着 Internet的兴起而产生的一种网络学术资源 ,近年来在国外发展较快。本文阐述了 e- print的概念、特点 ,以及 e- print发展存在的问题 ,并介绍了 arxiv e- print archive数据库及其检索方法。

Applying Neural-Network-Based Machine Learning to Additive Manufacturing:Current Applications,Challenges,and Future Perspectives

Additive manufacturing(AM),also known as three-dimensional printing,is gaining increasing attention from academia and industry due to the unique advantages it has in comparison with traditional subtractive manufacturing.However,AM processing parameters are difficult to tune,since they can exert a huge impact on the printed microstructure and on the performance of the subsequent products.It is a difficult task to build a process-structure-property-performance(PSPP)relationship for AM using traditional numerical and analytical models.Today,the machine learning(ML)method has been demonstrated to be a valid way to perform complex pattern recognition and regression analysis without an explicit need to construct and solve the underlying physical models.Among ML algorithms,the neural network(NN)is the most widely used model due to the large dataset that is currently available,strong computational power,and sophisticated algorithm architecture.This paper overviews the progress of applying the NN algorithm to several aspects of the AM whole chain,including model design,in situ monitoring,and quality evaluation.Current challenges in applying NNs to AM and potential solutions for these problems are then outlined.Finally,future trends are proposed in order to provide an overall discussion of this interdisciplinary area.

Fully Connected Convolutional Neural Network in PCB Soldering Point Inspection

In Electronics Manufacturing Services (EMS) industry, Printed Circuit Board (PCB) inspection is tricky and hard, especially for soldering point inspection due to the extremely tiny size and inconsistent appearance for uneven heating in reflow soldering process. Conventional computer vision technique based on OpenCV or Halcon usually cause false positive call for originally good soldering point on PCB because OpenCV or Halcon use the pre-defined threshold in color proportion for deciding whether the specific soldering point is OK or NG (not good). However, soldering point forms are various after heating in reflow soldering process. This paper puts forward a VGG structure deep convolutional neural network, which is named SolderNet for processing soldering point after reflow heating process to effectively inspect soldering point status, reduce omission rate and error rate, and increase first pass rate. SolderNet consists of 11 hidden convolution layers and 3 densely connected layers. Accuracy reports are divided into OK point recognition and NG point recognition. For OK soldering point recognition, 92% is achieved. For NG soldering point recognition, 99% is achieved. The dataset is collected from KAGA Co. Ltd Plant in Suzhou. First pass rate at KAGA plant is increased from 25% to 80% in general.

Design and Analysis of a Compact Band Notch UWB Antenna for Body Area Network

This paper presents the design of a small printed ultra wideband antenna with Band Notched characteristics. Both the free space and on-body performances of this antenna were investigated through simulation. The newly designed UWB antenna is more revised small form factor sized, with the ability to avoid interference caused by WLAN (5.15 - 5.825 GHz) and WiMAX (5.25 - 5.85 GHz) systems with a band notch. The return loss response, gain, radiation pattern on free space of the antenna were investigated. After that, the on-body performances were tested on 3-layer human body model with radiation pattern, gain, return loss, and efficiency at 3.5, 5.7, 8, 10 GHz and all the results were compared with free space results. As the on-body performance was very good, the proposed antenna will be suitable to be used for multi-purpose medical applications and sports performance monitoring.

基于神经网络的印染废水处理组合工艺系统构建

印染废水含有多种复杂的有机物质和重金属,处理难度较大,传统的处理技术往往难以达到理想的处理效果。近年来,人工智能技术为高效、精准的废水处理提供了新的解决方案,其中,神经网络以其强大的数据处理能力和模式识别功能可以优化处理工艺,提高处理效率,并降低运营成本。基于神经网络技术的印染废水处理组合工艺系统进行深入探讨,以期为印染行业提供一种更为高效、可持续的废水处理新方案,促进环保技术的应用和环境质量的改善。

The Era of Industry 4.0:Opportunities for Printing Industry

The era of Industry 4.0 is around the corner.In this era that is full of keywords such as automation,intelligence,networking,informatizatio the problem of how printing com panies can betterintegrate in this era,and seize the opportunities to devel op is very real.Since the processing information of printing industry has a

基于CycleGAN算法的凤翔木版年画色彩特征迁移研究

目的为传承保护和创新发展陕西地区非物质文化遗产——凤翔木版年画,解决传统手工艺色彩复原难、图案组合创新效率低等问题,通过生成对抗网络与凤翔木版年画结合的方式,多维度推动其发展。方法通过高清扫描、自适应阈值检测法、标准色值提取及数据增广等方式对其进行系统性归纳总结与标准化采集存储,建立凤翔木版年画数据集,利用CycleGAN算法训练网络模型以完成迁移实验。结果通过艺术与技术的结合,不仅完成了年画墨线稿的色彩复原,还完成了创新设计图案的快速着色。结论在凤翔木版年画的史料复原与设计应用方面,通过现代设计方法与色彩风格迁移功能的结合,使其在适应现代发展趋势、完成活态传承任务的同时,为同类研究提供新思路。

Multi-Objective Optimization Design through Machine Learning for Drop-on-Demand Bioprinting

Drop-on-demand (DOD) bioprinting has been widely used in tissue engineering due to its highthroughput efficiency and cost effectiveness. However, this type of bioprinting involves challenges such as satellite generation, too-large droplet generation, and too-low droplet speed. These challenges reduce the stability and precision of DOD printing, disorder cell arrays, and hence generate further structural errors. In this paper, a multi-objective optimization (MOO) design method for DOD printing parameters through fully connected neural networks (FCNNs) is proposed in order to solve these challenges. The MOO problem comprises two objective functions: to develop the satellite formation model with FCNNs;and to decrease droplet diameter and increase droplet speed. A hybrid multi-subgradient descent bundle method with an adaptive learning rate algorithm (HMSGDBA), which combines the multisubgradient descent bundle (MSGDB) method with Adam algorithm, is introduced in order to search for the Pareto-optimal set for the MOO problem. The superiority of HMSGDBA is demonstrated through comparative studies with the MSGDB method. The experimental results show that a single droplet can be printed stably and the droplet speed can be increased from 0.88 to 2.08 m·s^-1 after optimization with the proposed method. The proposed method can improve both printing precision and stability, and is useful in realizing precise cell arrays and complex biological functions. Furthermore, it can be used to obtain guidelines for the setup of cell-printing experimental platforms.

馆藏绝版作品数字化传播的规范进路

国家图书馆与北京三面向公司著作权纠纷案让馆藏绝版作品的数字化传播问题引起社会关注。为保护著作权并调和社会公共利益,避免破坏正常市场运作机制,我国《信息网络传播权保护条例》规定了馆藏绝版作品数字化传播的合理使用制度。从实施情况看,条例对适格客体条件与传播范围的限制过于严苛,存在违反基本公共服务均等性要求和违背利益平衡原则等问题。借鉴新修订《日本著作权法》中的相关条款,建议我国对于馆藏绝版作品的数字化传播规范宜采取如下改革措施:一是扩大合理使用的范围,建立覆盖馆内、馆际、馆外的传播体系,并设置匹配的补偿机制;二是优化客体适用条件,建立“绝版作品”的动态认定机制;三是探索建立以国家图书馆为枢纽的绝版作品数字化传播运行机制。

基于三重注意力的轻量级YOLOv8印刷电路板缺陷检测算法

在全球产业中,印刷电路板的生产和应用持续增长,已经成为各种电子设备的核心组成部分。由于缺陷尺度较小的问题以及检测模型轻便嵌入便携式设备的需求,印刷电路板图像的自动缺陷检测是一项具有挑战性的任务。为了满足智能制造和使用中对高质量印刷电路板产品日益增长的需求,提出一种基于YOLOv8的印刷电路板缺陷检测改进方法。首先,采用轻量级网络MobileViT作为主干网络,减小模型体积和计算量。其次,引入Triplet Attention模块,增强张量中不同维度间特征的捕捉能力。最后,将边界框损失函数替换为LMPDIoU,直接最小化预测框与实际标注框之间的左上角和右下角点距离。实验表明:改进后的检测模型能够在拥有极小参数量的同时保证小尺寸缺陷检测精度较高,模型参数量降低率为89.38%,满足轻便嵌入便携式检测设备和计算机资源受限的场景应用,证实了在印刷电路板缺陷检测领域具有良好的应用前景。

基于回归分析和GA-BP神经网络算法的3D打印件弯曲性能预测

为进一步探究熔融沉积成型(FDM)3D打印参数和制件弯曲性能之间的关系,创建合理的FDM 3D打印制件弯曲强度预测模型。根据正交试验L_(16)(4^(5))的设计原则和神经网络算法模型的构建要求,按照不同分层高度、填充密度、打印温度、打印速度以及外壳厚度五种因素,制备25组试验试样,并进行弯曲性能检测。随后通过建立GA-BP神经网络模型、传统BP神经网络模型以及多元回归方程模型,分别对FDM 3D打印制件弯曲性能进行预测,并将预测数据与试验测试数据进行对比。通过对比发现,GA-BP神经网络模型预测数据与试验测试数据更为接近,其平均误差为3.71%,且误差值整体波动最小,BP神经网络模型与多元回归方程模型预测精度相差不大,BP神经网络模型预测平均误差为8.05%,多元回归方程模型预测平均误差为9.07%,但多元回归方程误差值整体波动最大。因此,采用GA遗传算法优化后的BP神经网络模型在进行FDM 3D打印制件弯曲性能预测方面具有更高的精度和更良好的稳定性。

基于神经网络的拓扑与打印方向同步优化

三维打印能够制造复杂的结构,并广泛应用于航空,工业,医学等诸多行业,但它也存在着一些局限性.而拓扑优化方法是在空间内寻求最佳的材料分布,从而使用更少的材料实现更轻的部件与同样的性能.近年来,三维打印结合拓扑优化的方法受到了广泛关注.众所周知,零件的打印方向会影响打印效果,并与拓扑结构,支撑体积等方面具有密切关系.鉴于此,文中提出了一种基于神经网络的打印方向优化框架.为了验证算法的可行性,根据结构柔度与所需支撑体积构造可微的损失函数,并使用人工神经网络同时对结构拓扑与打印方向进行优化.通过在二维和三维上的大量实验证明,方向优化可以在保证柔度变化不大的情况下显著减少支撑体积数量,降低打印成本.

人工神经元网络模型预测3D打印部件力学性能的研究

熔融沉积成型(FDM)是一种高效的增材制造技术。将响应面模型与人工神经元网络(ANN)模型相结合,研究了FDM工艺的喷嘴温度、层高和层积角度对尼龙12(PA12)丝材制造部件力学性能的影响。当喷嘴温度、层高和层积角度分别在220~260℃、0.2~0.4 mm、0°~90°之间变化时,部件拉伸强度和缺口冲击强度分别在35.69~60.89 MPa和5.48~19.83 kJ/m^(2)之间。喷嘴温度、层高、层积角度以及层积角度的二阶效应是影响部件拉伸强度的显著因素;喷嘴温度、层积角度以及层积角度的二阶效应是影响缺口冲击强度的显著因素。ANN模型预测拉伸强度和缺口冲击强度的最优结构分别是3-10-5-1和3-25-24-1,预测的拉伸强度和缺口冲击强度均方误差函数(MSE)最低分别为2.54×10^(-4)和2.07×10^(-4),回归系数均在0.97以上。与响应面的二次回归模型相比,ANN模型预测的拉伸强度和缺口冲击强度与实验值的标准偏差分别为0.46和0.32,远低于二次回归模型的2.43和1.58,更适合于优化非线性的FDM工艺。

Web Tension Regulation of Multispan Roll-to-Roll System using Integrated Active Dancer and Load Cells for Printed Electronics Applications

The mass production of primed electronics can be achieved by roll-to-roll（R2R） printing system, so highly accurate web tension is required that can minimize the register error and keep the thickness and roughness of printed devices in limits. The web tension of a R2R system is regulated by the use of integrated load cells and active dancer system for printed electronics applications using decentralized multi-input-single-output（MISO） regularized variable learning rate backpropagation artificial neural networks. The active dancer system is used before printing system to reduce disturbances in the web tension of process span. The classical PID control result in tension spikes with the change in roll diameter of winder and unwinder rolls. The presence of dancer in R2R system shows that improved web tension control in printing span and the web tension can be enhanced from 3.75 N to 4.75 N. The overshoot of system is less than ±2.5 N and steady state error is within ± 1 N where load cells have a signal noise of ±0.7 N. The integration of load cells and active dancer with self-adapting neural network control provide a solution to the web tension control of multispan roll-to-roll system.

融合自注意力与残差神经网络的3D打印激光在机测量误差修正方法

激光测量能够实现高效地非接触实时测量,被广泛应用于3D打印领域,但激光测量容易受测量条件、外部环境等多种因素的干扰,这些因素错综复杂,难以量化分析。为此,结合直射式激光三角测量原理,在分析测量精度影响因素的基础上,提出了一种基于融合自注意力和残差神经网络的3D打印在机测量误差修正方法。首先,将影响测量精度的因素作为输入变量,采集激光测量值,得到样本数据集;然后利用残差网络提取出样本数据的深层次特征,并引入自注意力机制建立影响因素之间的联系,得到带权重的提取特征;再通过全连接网络对带权重特征进行学习,得到测量误差的预测值,基于该预测值完成对测量误差的修正。自主搭建了一套激光在机测量系统,采用红、绿、紫3种同材质彩色卡纸进行实验验证。结果表明,所提的方法与卷积神经网络和自注意力神经网络相比,均方误差、均方根误差和平均绝对误差均最小,稳定性最好,修正结果最接近真实值;对激光测量结果进行校正后,使其误差由原来的±28μm减小到±9μm以下,显著提高了3D打印激光在机测量的精度和稳定性。

包装印刷设备网络协同制造模式实施路径研究

推行包装印刷设备网络协同制造模式是促进产业转型和实现高质量发展的重要方法。为给企业实施网络协同制造模式提供操作框架和实践参考,本研究对包装印刷设备产业的现状和特征进行分析,基于现实案例,参考精益屋模型,提出了包装印刷设备网络协同制造模式的框架模型,并对包装印刷设备网络协同制造模式的实施所面临的困难和挑战进行分析。

神经网络技术在印染配色中的应用

随着纺织印染工业对配色精度和效率的要求不断提高,传统的配色技术主要依赖于设计师的主观判断,已经不能满足大批量和高变异性的生产要求。随着人工智能技术的发展,神经网络在印染配色中的应用可以模拟人类视觉系统的处理机制,通过训练识别不同颜色和图案,自动进行色彩匹配和调整。探讨基于神经网络技术的印染配色模型,建立和集成基于神经网络技术的印染配色系统,并通过仿真实验验证模型的有效性,以期为纺织印染行业提供可行的印染配色技术方案参考。

基于BP神经网络的气溶胶喷射3D打印技术质量建模及分析

为了提高气溶胶喷射3D打印质量的稳定性和准确性,建立基于Back Propagation(BP)神经网络的打印质量预测模型.该模型以鞘气流量、载气流量和打印速度为主要参数,并预测气溶胶喷印中的两个重要指标:线条宽度和线条粗糙度.同时,采用了K折交叉验证方法对神经网络模型进行训练,并对网络结构进行了评估.测试结果表明,该模型具有较高的预测精度和稳定性,能够准确地预测线条宽度和线条粗糙度.