用于机器人关节的柔性压力触觉传感器设计

分析了碳纤维/硅橡胶导电复合材料的拉伸、弯曲及压敏特性,基于上述特性设计了一种新型的应用于机器人关节等活动部位的柔性压力触觉传感器。通过解耦算法解决了拉伸、弯曲引起的干扰问题,构建了求解接触压力的数学模型。实验结果表明,该柔性触觉传感器具有拉伸和弯曲性,并消除了拉伸和弯曲对接触压力的检测的干扰,可应用于机器人关节。     

汽车排气系统吊耳动刚度优化方法的研究

基于汽车排气系统吊耳传递的动态载荷最小、吊耳耐疲劳性最好,建立了考虑动力总成在内的排气系统振动分析模型。进行了排气系统的自由模态和约束模态的测试,并和计算值进行了对比分析,证明了所建立的排气系统振动模型的正确性。以吊耳的垂向动态载荷最小和其静变形量在一定范围内为优化目标,建立了排气系统吊耳动刚度优化模型。优化后,在怠速工况和2档全负荷加速工况下对车身底板驾驶员位置进行了振动响应测试,测试结果表明,利用优化后的吊耳刚度,能够有效降低车身底板的振动加速度,表明了阐述的排气系统建模和吊耳动刚度优化方法的有效性。文中建模与优化方法,对排气系统的吊耳动刚度计算与优化具有指导意义。     

Modeling of multicomponent diffusions and natural convection in unfractured and fractured media by discontinuous Galerkin and mixed methods

Computation of the distribution of species in hydrocarbon reservoirs from diffusions (thermal, molecular, and pressure) and natural convection is an important step in reservoir initialization. Current methods, which are mainly based on the conventional finite‐difference approach, may not be numerically efficient in fractured and other media with complex heterogeneities. In this work, the discontinuous Galerkin (DG) method combined with the mixed finite element (MFE) method is used for the calculation of compositional variation in fractured hydrocarbon reservoirs. The use of unstructured gridding allows efficient computations for fractured media when the cross flow equilibrium concept is invoked. The DG method has less numerical dispersion than the upwind finite‐difference methods. The MFE method ensures continuity of fluxes at the interface of the grid elements. We also use the local DG (LDG) method instead of the MFE to calculate the diffusion fluxes. Results from several numerical examples are presented to demonstrate the efficiency, robustness, and accuracy of the model. Various features of convection and diffusion in homogeneous, layered, and fractured media are also discussed.     

形态发生学视角下的家居器皿参数化设计研究

目的 探索形态发生学视角下的参数化设计策略及应用,为设计领域创新提供新的视角和可能性,并为设计教育和设计实践的学科交叉融合提供研究案例。方法 运用形态发生学原理和参数化设计模型生成,以家居桌面器物为设计对象展开实验性设计,并利用数字制造技术完成设计方案的实物化。结果 形态发生学视角下参数化设计策略的探索取得显著进展,并在家居器皿设计实践中得到成功验证,取得创新性成果。结论 不同学科知识和方法的交叉融合能够促进形态发生原理的理解和应用;数字技术与自然模式的协同、计算力学的综合应用,以及数学模型的设计转化和集成,能够大大增强形态生成的精确控制力,并带来更大的创新空间,是设计领域创新的科学路径之一。     

基于深度学习的孔特征可制造性分析方法

针对传统基于知识库及规则库的零件可制造性分析方法柔性差,以及现有基于深度学习的可制造性分析方法无法给出零件具体不可制造原因的现状,提出一种基于深度学习的零件可制造性分析方法.首先,通过数字化建模技术构建大量带有具体可制造性类别标签的三维CAD模型,并进行点云提取,从而构建深度学习所需数据集;然后,基于PointNet网...     

基于带约束三角剖分的三维巷道建模方法

提出了具有可行性的巷道模型实现方法,将巷道模型抽象为顶盘面、底盘面和侧面3部分。在获取巷道顶盘（或底盘）边界测量数据的条件下,提出了采用约束三角剖分的方法生成巷道模型的上下底盘面,利用两段成面法生成巷道模型的侧面,最后将这些面组合成巷道模型。该方法可以一次性地生成某一开采层面的全部巷道模型,并已成功地运用到以GeoView为平台的“数字紫金”项目中,取得了良好的效果。     

采用磁流变阻尼器的汽车悬架系统的杂交建模

利用线性回归方法,对磁流变阻尼器阻尼力与速度、加速度间的函数关系作了线性与非线性的区分,采用神经网络对非线性部分进行建模,并结合线性部分构建了磁流变阻尼器的模型。在此基础上,融合汽车悬架的先验知识模型,建立了采用磁流变阻尼器的4自由度1/2车辆半主动悬架系统的杂交模型。最后利用SANTANA 2000型轿车的参数进行仿真,并与基于磁流变阻尼器非线性滞回模型的建模方法作了比较,结果表明:杂交建模方式结构简单、计算量小、模型准确、便于进行系统主要动力学特性的分析。     

Improvement of lubrication and cooling in grinding

The presence of air boundary layer around a grinding wheel impedes the proper entry of cutting fluid into grinding zone and thereby lubrication and cooling are hampered. It leads to the thermal damage to the workpiece and rapid wheel wear. This experimental work is aimed at increasing the lubrication and cooling effects in grinding by a novel method of using scraper board. The experiments are conducted, using the scraper board, to find the critical region where the air pressure is zero. The coolant jet has then impinged into the grinding zone which is maintained within the critical region. The requirement of specific energy, surface texture, and mechanical properties of the ground surface are analyzed and compared with the traditional grinding and the grinding at various positions of scraper board. Results show that the grinding ratio improved by 35.6and 119%, surface roughness decreased by 36 and 54.2%, while the requirement of specific energy is reduced by 50.8 and 57.3% when scraper board is positioned at the critical distance in comparison to the 57.5° position and no scraper board, respectively. The results indicate that the introduction of fluid in the present method can improve the process efficiency and the product quality effectively.     

角域AR谱技术在齿轮故障诊断中的应用

利用时频分布平面内信号能量峰脊与瞬时频率之间的对应关系,对信号瞬时频率进行估计;在此基础上利用代数方法求解鉴相时标积分方程,并对经插值重采样得到的角域信号进角域平均处理,提高了角域信号的信噪比;最后对角域信号进行AR建模实现信号的阶次谱分析。实际测试结果表明:采用角域AR谱技术处理齿轮箱非平稳振动信号,能够有效地避免传统频谱方法无法解决的"频率模糊"现象,克服了传统阶次谱分辨率较低,谱线毛糙,易受噪声及轴频调制影响等缺点,对齿轮箱的早期故障有较好的识别能力。     

一种弧长参数曲线及其计算机实现

本文提出了一种弧长参数曲线的数值算法，这种算法具有收敛迅速，迭代过程自动确定等优点。