基于深度剥离的三维打印模型朝向优化算法

为了提高三维打印产品表面质量且有效节省材料,提出一种面向熔融挤压式三维打印机的模型朝向优化算法.该算法基于深度剥离技术,以打印朝向为自变量构建模型悬空面积及其支撑结构体积的双目标优化函数;利用模式搜索方法设计一个两步走的优化方案来确定最优的打印朝向.实验结果表明,利用深度剥离技术能够较为准确、高效地检测出打印模型悬空区域,并计算出相应的悬空面积和支撑体积;设计的优化方案也能够根据双目标的权重比确定一个较优的打印朝向,不但减少了支撑结构与模型表面的接触范围,还能降低支撑结构的体积.文中算法在提高模型表面质量以及减少支撑结构的材料消耗两方面具有优势.     

基于体素模型的3D打印支撑算法

3D 打印技术是通过对材料的逐层堆积来构建物体,但对模型悬空的区域需要添加 支撑结构。支撑结构不仅会造成打印材料的浪费,而且会延长打印时间并对模型外表有所损伤。 为此,提出一种基于体素模型的支撑算法,针对体素化后的模型,分析体素之间的相互支撑作用, 并引入体素支撑能量函数概念和计算方法,计算出需要添加支撑的体素,从而得到需要添加支撑 结构的区域,并由该区域生成支撑结构,之后通过实验对算法进行验证。实验结果显示该算法能 够更加准确地对模型生成支撑,同时,基于体素模型的支撑算法对于模型内部支撑计算,也具有 很好的适用性。     

装配体模型优化的免装配打印方法EI北大核心CSCD

针对包含机构语义的装配体模型实施高质量的免装配打印(涉及支撑结构的清除、打印材料的消耗以及机构语义关联表面的打印质量等)不仅困难且仍缺乏自动方法的问题,提出一种装配体模型优化的免装配打印方法.首先基于包围盒剖分和体素关联关系对装配体模型进行体素化;然后采用改进的Bresenham算法快速生成体素路径,并据此快速分析支撑结构的可清除性;提出一种增量式的体素重关联方法,可在零件之间的相对布局改变后避免耗时的模型重体素化;最后以同时满足支撑结构可清除、支撑结构消耗材料量尽量少,以及运动接口面的打印质量尽量好为目标,通过改进粒子群算法确定和调整装配体模型的打印方向及其零件之间的相对布局.在带有典型机构语义的若干装配体模型上进行实验的结果表明,所提方法计算的方向和零件之间的相对布局打印模型的支撑结构均可完全清除,支撑结构所消耗的材料以及机构语义关联表面的粗糙度均大幅度减少(小);与基于八叉树分析支撑结构可清除性的方法以及基于零件布局调整开展免装配机构打印的最新方法相比,该方法对于包含机构语义的装配体模型开展高质量免装配打印具有突出的适用性和优势.     

FDM快速成型支撑结构自动生成算法的研究

FDM系统中零件原形的加工是靠热喷头挤出熔丝逐层堆积而成的。因此,支撑结构的添加对零件原形的制作起着至关重要的作用。针对支撑结构的需求,采用了一种新的支撑自动生成算法。该算法基于STL模型首先需提取待支撑区域,然后添加相应的支撑结构,最后确定待支撑区域的起点和终点。实际应用表明,生成支撑的效果良好。     

基于高斯球的近似共面合并层次细节模型

构造场景多层次细节模型是加速图形显示的有效途径.给出一种基于共面合并自动生成场景多层次细节模型的算法.此算法将一单位球定义为高斯球,并按一定的规则对其进行层次分割.然后根据各景物面片的法向在高斯球上的分布,将具有近似一致法向的面片划分为面片族;在面片族内对相邻面片进行合并,生成近似共面区域;然后移去其内部点,对区域边界重新三角化.算法还结合临近边界点合并等原则,删除对表达场景几何特征重要度低的顶点.算法利用平面九叉树进行合并操作,生成具有多层次细节的场景模型.实验结果表明,算法能实现较理想的场景简化效果.     

面向耗材节省的三维打印路径规划算法研究

为降低三维打印（three-dimensional printing,3D）耗材费用并进一步提高打印效率,给出一种面向熔融沉积制造的三维打印路径规划算法。该方法综合考虑打印耗材、打印效率以及打印表面质量等因素,通过网格模型及其支撑的相邻层片轮廓关系求得可稀疏打印区域;基于多边形扫描线算法以及多边形单调链关系,得到能够连续打印的路径区域;最终通过区域路径稀疏化得到改进的打印路径。通过复杂网格模型的三维打印路径规划实例,验证了算法的有效性。该算法能够降低打印耗材数量,并进一步提高打印效率。     

NURBS体参数化模型的无支撑打印算法

处理大角度悬垂结构的打印问题是无支撑打印的主要挑战。针对NURBS（non-uniform rational Bspline）体参数化模型的多自由度无支撑三维打印技术,提出一种锥形切片算法。通过对模型的三维文件进行几何映射,利用水平切片算法对其进行切片,得到模型变形后的连续打印路径G-code;然后对得到的G-code进行逆映射,生成适用于六轴机械臂的多自由度连续打印路径。利用锥形切片算法规划打印路径,可以在无须支撑的情况下,依靠自支撑结构实现对悬空结构部分的打印。最后,通过仿真实验和对比实验,验证了锥形切片算法的准确性、可行性和有效性。锥形切片算法的引入为NURBS体参数化模型的无支撑三维打印提供了一种更加高效、经济和环保的解决方案。     

内部结构自支撑的三维打印浮体平衡优化算法

为了使浮体能够在熔融挤压式三维打印机中一次成型,提出一种内部结构自支撑的浮体平衡优化算法.首先在给定漂浮姿势的输入模型内部,利用递归细分方法生成自支撑的格子结构,并把输入模型与格子结构组合为初始浮体模型.然后以初始浮体模型中格子单元体积为变量构建平衡优化方程,根据方程解算结果调整初始浮体模型中每个格子单元的体积,从而生成最终的浮体模型.实验结果表明,文中算法生成的浮体模型可以使用熔融挤压式三维打印机直接打印,打印的实体不但能够按照预设的姿势漂浮在液体中,而且具有较好的抗压能力;算法中的递归细分策略能够在模型内部产生较多的格子单元,具有节省打印材料的优势.     

面向3D打印的自支撑连通性填充结构设计

针对大多数3D打印轻质化填充结构因不能够自支撑而导致模型无法打印成型,额外添加支撑结构又会造成模型内部残留打印材料无法排出的问题,提出一种面向3D打印的自支撑连通性填充结构设计方法.首先,基于每层打印切片构建填充结构;其次,通过调整参数来控制不同层正三角形连续地周期性变化,实现填充结构的自支撑性和连通性;最后,根据每层的几何信息生成可以直接用于打印的G-code文件.通过进行打印实物、打孔灌水和力学测试实验,实验结果表明,该方法可以打印具有不同自支撑角度且包含2个连通区域的填充结构,与5种典型填充结构相比,该填充结构力学性能良好.     

半边结构的爆炸破片生成方法

研究虚拟实体爆炸过程中破片的生成是模拟爆炸真实性及杀伤力的关键和难点，在深入探讨爆炸模型的构造和实时显示的基础上，提出隐式破片、显式破片和随机半边结构的最新概念，以及两种生成破片的方法：小面片等价法和面片集封闭法。给出三个实用的算法：虚拟实体的微粒化算法、小面片等价算法和面片集封闭算法；这三个算法的复杂度均为O（n）。最后，对这两种生成破片的方法及其算法进行了比较；对这两种方法在实体破裂过程中的运用进行了探讨；同时对实际运行的时间进行了测量和对照；并通过了两个实例的验证。     

熔丝沉积制造中稳固低耗支撑结构生成

熔丝沉积制造(Fused deposition modeling, FDM)是利用熔融塑料丝的一种3D打印技术,热塑料由喷嘴喷出逐层堆积完成打印.由于熔丝只能沉积在已存在物体的上层,因此需要构造支撑结构以支撑悬空部分.针对现有支撑结构生成算法中存在的或结构不稳固或耗材多的缺陷,提出一种以熔丝为支撑单位的树形稀疏支撑结构.与传统算法计算模型表面支撑区域不同,本算法计算每段熔丝需要支撑的区域,使支撑结构更契合熔丝沉积特点.算法还将支撑结构分为三类,将多约束优化问题分解,降低算法复杂度.实验结果表明,本文算法生成的支撑结构算法耗材少、支撑稳定.     

Attribute driven process architecture for additive manufacturing

In additive manufacturing (AM) process, the manufacturing attributes are highly dependent upon the execution of hierarchical plan. Among them, material deposition plan can frequently interrupt the AM process due to tool-path changes, tool start-stop and non-deposition time, which can be challenging during free-form part fabrication. In this paper, the layer geometries for both model and support structure are analyzed to identify the features that create change in deposition modality. First, the overhanging points on the part surface are identified using the normal vector direction of the model surface. A k-th nearest point algorithm is implemented to generate the 3d boundary support contour which is used to construct the support structure. Both model and support structures are sliced and contours are evaluated. The layer contour, plurality, concavity, number of contours, geometric shape, size and interior islands are considered to generate an AM deposition model. The proposed model is solved for minimizing the change in deposition modality by maximizing the continuity and connectivity in the material deposition plan. Both continuity and connectivity algorithms are implemented for model and support structure for free-form object. The proposed algorithm provides the optimum deposition direction that results in minimum number of tool-path segments and their connectivity while minimizing contour plurality effect. This information is stored as a generic digital file format named Part Attributable Motion (PAM). A common application program interface (API) platform is also proposed in this paper, which can access the PAM and generate machine readable file for different existing 3D printers. The proposed research is implemented on three free-form objects with complex geometry and parts are fabricated. Also, the build time is evaluated and the results are compared with the available 3d printing software.     

DMS2015-33: Generative interface structure design for supporting existing objects

Increasing availability of high quality 3D printing devices and services now enable ordinary people to create, edit and repair products for their custom needs. However, an effective use of current 3D modeling and design software is still a challenge for most novice users. In this work, we introduce a new computational method to automatically generate an organic interface structure that allows existing objects to be statically supported within a prescribed physical environment. Taking the digital model of the environment and a set of points that the generated structure should touch as an input, our biologically inspired growth algorithm automatically produces a support structure that when physically fabricated helps keep the target object in the desired position and orientation. The proposed growth algorithm uses an attractor based form generation process based on the space colonization algorithm and introduces a novel target attractor concept. Moreover, obstacle avoidance, symmetrical growth, smoothing and sketch modification techniques have been developed to adapt the nature inspired growth algorithm into a design tool that is interactive with the design space. We present the details of our technique and illustrate its use on a collection of examples from different categories.     

A development of data structure and mesh generation algorithm for whole ship analysis modeling system

In the whole ship structure and vibration analysis, the FEA (finite element analysis) model of whole ship structure is required in the early design stage before the 3D CAD model is defined. Because ship structure has a complex curved surface, and many associated structural members, the whole ship analysis modeling job has become a time consuming job. For the effective support of the whole ship analysis modeling, a method to generate the analysis model using initial design information within the ship design process, hull form offset data and compartment data, is developed. To easily handle initial design information and FE model information, a flexible data structure is proposed. An automatic quadrilateral mesh generation algorithm using initial design information to satisfy the constraints imposed by the ship structure is also proposed. With the proposed data structure and mesh generation algorithm, whole ship analysis modeling job for various ship types can be effectively supported and these results are presented.     

基于深度SSD改进模型的传动设备状态在线监测研究

针对现有传动设备在线监测算法存在的检测精度地、效率差等问题,提出一种基于改进SSD网络模型的在线检测算法。先对故障集进行预处理,通过滤波调制、共振解调等环节滤除原始故障集的噪声干扰;以VGG-16为基础设计了SSD网络结构,同时增加了辅助卷积层和预测层;对SSD网络模型进行改进,引入了注意力机制模块和特征增强模块,改善模型各层的数据共享性能同时提高了模型的数据训练效率;基于通道拼合方式对故障数据进行多尺度特征融合,并优化SSD模型的各层金字塔结构,以更好的匹配先验框及选择最佳的损失函数。实验结果显示,提出算法的传动设备故障检测率达到98.8%,同时算法的检测效率也优于现有算法。     

一种基于逻辑栅格的网格自动生成算法

针对基于栅格法对表面变动较大的模型进行网格划分适应性差的问题,提出一种基于逻辑栅格的六面体网格自动生成算法。该算法对基于栅格法的思想进行改进,避开一次生成模型初始栅格,将模型离散为一系列截面,对相邻截面采用矢量求交方法形成一对截面的初始栅格,修正初始栅格形成两截面之间的一层网格,形成整体模型。采用改进的栅格法进行复杂地形拱坝坝肩块体网格自动生成,取得较好的效果,结果证明改进的栅格法对复杂地形适应性强,模型表面变动较大时能自动进行网格退化。     

Ribbed Support Vaults for 3D Printing of Hollowed Objects

Additive manufacturing techniques form an object by accumulating layers of material on top of one another. Each layer has to be supported by the one below for the fabrication process to succeed. To reduce print time and material usage, especially in the context of prototyping, it is often desirable to fabricate hollow objects. This exacerbates the requirement of support between consecutive layers: standard hollowing produces surfaces in overhang that cannot be directly fabricated anymore. Therefore, these surfaces require internal support structures. These are similar to external supports for overhangs, with the key difference that internal supports remain invisible within the object after fabrication. A fundamental challenge is to generate structures that provide a dense support while using little material. In this paper, we propose a novel type of support inspired by rib structures. Our approach guarantees that any point in a layer is supported by a point below, within a given threshold distance. Despite providing strong guarantees for printability, our supports remain lightweight and reliable to print. We propose a greedy support generation algorithm that creates compact hierarchies of rib-like walls. The walls are progressively eroded away and straightened, eventually merging with the interior object walls. We demonstrate our technique on a variety of models and provide performance figures in the context of fused filament fabrication 3D printing.     

基于应力分布的3D打印壳模型构造与优化

近年来,愈发成熟的3D打印技术拉近了模型设计与产品制造的距离.但高昂的材料费用仍是制约其发展的重要因素.因此,如何在不改变模型外观的情况下进行模型结构的优化,以此来减少模型的打印体积、降低打印成本是亟待解决的问题.针对该问题,本文提出一种基于应力分布的壳模型构造和优化算法.该算法首先基于模型的体素化表示构造距离场,提取出初始的均匀厚度壳模型.然后基于顶点的von Mises应力,自适应的向外扩张内表面,调整各部位厚度,直到达到相关约束条件.优化得到的内表面与输入的外表面围成最终的优化模型.实验结果表明,在满足外观不变、力学稳定等约束的同时,优化得到的壳模型体积为输入模型体积的17.2%~24.4%,大幅减少了模型的打印体积,有效降低了打印成本.     

基于深度学习的航空发动机传感器故障检测

针对传统反向传播(BP)神经网络和支持向量机(SVM)存在的过拟合、维数灾难、参数选择困难等问题,提出了一种基于深度学习算法的航空发动机传感器故障检测方法.对发动机参数记录仪采集的多维数据进行预处理,建立基于深度置信网络(DBN)的故障检测模型,利用预处理后的数据对检测模型进行训练,经过DBN故障检测模型逐层特征学习实现了传感器故障检测.仿真结果表明:在无人工特征提取和人工特征提取的情况下,基于DBN故障检测的准确率均高于BP神经网络和SVM模型.