稀土变质高洁净轴承钢中夹杂物的行为分析

为了阐明稀土变质高洁净轴承钢中夹杂物的行为,进而优化轴承钢的稀土处理过程,通过向高洁净轴承钢中加入高纯稀土的方法,借助SEM和EDS等观察和分析手段,重点对不同稀土含量轴承钢中夹杂物类型和析出行为进行了表征和分析。结果表明,稀土能够变质高洁净轴承钢中的Al₂O₃和MnS而形成稀土夹杂物,RE-S-As-P夹杂物可以作为RE-O-S-As-P-C夹杂物的形核核心促进其析出。随稀土含量的增加,稀土元素与夹杂物的形成元素的结合序列依次为O、S、As、P和C,稀土夹杂物类型演化序列主要为RE₂O₃、RE₂O₂S、RES、RE-O-S-As、RE-S-As、RE-S-As-P、RE-O-S-As-P-C、RE-O-As-P-C、RE-O-P-C和RE-O-C。     

基于真实载荷的挖掘机工作装置瞬态动力学分析

用有限元法对挖掘机工作装置进行瞬态动力学分析,以6 t小型挖掘机工作装置为研究对象,针对斗杆液压缸驱动铲斗撞击地面工况,用压力传感器、位移传感器测试出撞击过程各液压缸工作腔压力和位移变化曲线,以所获各液压缸位移变化曲线和最大理论撞击力为驱动,用动力学仿真软件ADMAs对挖掘机撞击过程进行仿真,得出各铰销点在撞击过程中所承受载荷的变化曲线,采用测试所得各液压缸的驱动力验证仿真结果的准确程度.进一步将各铰销点受力的仿真结果作为工作装置的负载,对工作装置进行瞬态动力学分析,对比仿真计算与应力测试结果表明,对应测点的应力变化趋势基本一致,误差在10%以内,可为结构优化设计提供依据.     

滚动转子式压缩机容积效率计算模型

探讨了滚动转子式压缩机容积效率的模拟问题.针对目前纯理论模拟方法的复杂性及不实用性等局限,采用了试验曲线模拟的方法.通过对经典性能关联式进行修正建立起一系列优化的性能关联式,依据大量有效的实验数据,运用多元非线性拟合方法求解性能关联式系数.结果表明,由该方法所建立的模型能够准确地描述变工况下滚动转子式压缩机容积效率等性能参数.     

脉冲多弧离子源所镀膜层均匀性的实验研究

为了解决脉冲多弧离子源镀制膜层的均匀性问题,文中首先分析了不同阴极尺寸、不同辅助阳极尺寸、不同引弧方法以及不同基片距阴极的距离引起所镀膜层透过率曲线的改变,进而讨论了脉冲多弧离子源电极几何尺寸对膜厚均匀性的影响．采用在基片下加一个圆筒形负电位电极实现了脉冲多弧离子源镀膜的均匀性．实验结果表明,在直径７０ｍｍ范围内透过率相对误差为±６．７％,小于所要求的±１０％．     

工程车辆座椅减振环节模型探讨

分析和讨论了几种不同结构的座椅减振模型,并用某工程车辆的基本数据进行座椅的结构参数优化计算。结果表明,这几种结构模型均取得了较好的减振效果,即良好的乘坐舒适性,为工程车辆驾驶座椅的设计提供了有益的参考。     

识别油膜轴承动态特性参数的模态综合法

提出了一种现场在线识别转子系统油膜轴承动态特性参数的模态综合法,用广我特征值问题代替了大矩阵的求逆。本方法同传统的参数识别方法相比具有现场,在线,不需人工激振,设备简单,数据处理方便等特点。     

无菌条件非接触式多通道自然交互手术环境

无菌和非接触环境是医疗手术室的基本要求,这使得计算机操作室和手术室需要在物理上隔离.同时,因为手术进行中,主治医生如果需要查看病灶图像,通常授意护士或者手术助理到计算机操作室操作病灶图像,由于手术室和计算机操作室间的隔离,以及主治医生和助理间可能存在的意图理解不准确,容易导致护士或者手术助理在手术室和计算机操作室往返多次,这增加了患者手术时间延长、失血增多、脏器暴露时间长等风险,尽量减少手术中定位到病灶图像的时间对于医生和病人都很重要.针对上述需求,借助遮挡环境下的深度图像人体骨架提取、手势跟踪与理解、手术室环境远场语音识别,多模态信息处理与融合技术,构建无菌条件下的非接触式多通道自然交互手术环境.该环境使得主治医生在手术中可通过语音命令、手势及上述交互方式相结合的方式快速定位到需要观察的病灶成像.在接近真实环境的实验环境中,建立的无菌条件的非接触式多通道自然交互手术环境在保证精度的情况下,可显著缩短病灶图像定位时间.无菌环境智能交互医疗手术室为未来下一代高效的手术提供了技术与方法验证.     

An anti‐repeater‐jamming approach based on blind source separation for the downlink of satellite communication systems

In this paper, an anti‐jamming approach is proposed for the downlink of satellite communication systems when encountering a hostile repeater‐jamming. Based on blind source separation, this approach can eliminate repeater‐jamming by separating the mixtures of the communication signals and the repeater‐jamming. Meanwhile, oversampling method is employed to transform the underdetermined mixing of signals into a determined mixing for facilitating the separation. In the simulations, the symbol error ratio (SER) of the separated communication signals can approximate the theory SER, and the anti‐repeater‐jamming capacity can arrive to nearly 28 dB.     

Tribotronics for Active Mechanosensation and Self‐Powered Microsystems

Tribotronics has attracted great attention as a new research field that encompasses the control and tuning of semiconductor transport by triboelectricity. Here, tribotronics is reviewed in terms of active mechanosensation and human–machine interfacing. As a fundamental unit, contact electrification field‐effect transistors are analyzed, in which the triboelectric potential can be used to control electrical transport in semiconductors. Several tribotronic functional devices have been developed for active control and information sensing, which has demonstrated triboelectricity‐controlled electronics and established active mechanosensation for the external environment. In addition, the universal triboelectric power management strategy and the triboelectric nanogenerator‐based constant sources are also reviewed, in which triboelectricity can be managed by electronics in the reverse action. With the implantation of triboelectric power management modules, the harvested triboelectricity by various kinds of human kinetic and environmental mechanical energy can be effectively managed as a power supply for self‐powered microsystems. In terms of the research prospects for interactions between triboelectricity and semiconductors, tribotronics is expected to demonstrate significant impact and potential applications in micro‐electro‐mechanical systems/nano‐electro‐mechanical systems (MEMS/NEMS), flexible electronics, robotics, wireless sensor network, and Internet of Things.     

Hyperparameter Optimization for Machine Learning Models Based on Bayesian Optimization

Hyperparameters are important for machine learning algorithms since they directly control the behaviors of training algorithms and have a significant effect on the performance of machine learning models. Several techniques have been developed and successfully applied for certain application domains. However, this work demands professional knowledge and expert experience. And sometimes it has to resort to the brute-force search. Therefore, if an efficient hyperparameter optimization algorithm can be developed to optimize any given machine learning method, it will greatly improve the efficiency of machine learning. In this paper, we consider building the relationship between the performance of the machine learning models and their hyperparameters by Gaussian processes. In this way, the hyperparameter tuning problem can be abstracted as an optimization problem and Bayesian optimization is used to solve the problem. Bayesian optimization is based on the Bayesian theorem. It sets a prior over the optimization function and gathers the information from the previous sample to update the posterior of the optimization function. A utility function selects the next sample point to maximize the optimization function. Several experiments were conducted on standard test datasets. Experiment results show that the proposed method can find the best hyperparameters for the widely used machine learning models, such as the random forest algorithm and the neural networks, even multi-grained cascade forest under the consideration of time cost.