基于压缩感知理论重构压电阻抗信号的钢结构损伤识别

基于压电阻抗(electromechanical impedance, EMI)技术,针对结构损伤识别提出了阻抗/导纳数据压缩与重构的方法,利用随机矩阵将监测系统中的原始阻抗数据向量进行线性映射,并将映射后的向量输入到接收系统中;基于压缩感知理论将重构原始数据的问题转化成非确定性多项式问题,并基于凸优化(convex optimization, CO)理论求解;在损伤识别阶段,利用均方根偏差(root mean square deviation, RMSD)统计指标对重构阻抗数据的识别效果进行评估,并与使用原始阻抗数据的效果进行对比。利用简支钢梁的局部损伤识别试验采集的阻抗数据证明所提出方法的有效性。结果表明：基于重构阻抗数据能够有效识别结构损伤,基系数矩阵的稀疏度随着测量数的减少而降低,一致球集合对应的稀疏度区间低于其他测量矩阵,阻抗数据重构效果随着压缩率的增加而减弱,当压缩率高于2.0时,部分使用重构阻抗数据识别结构损伤的误差将大于20%,损伤识别精度降低。     

Robust stability analysis and improved design of phase‐locked loops with non‐monotonic nonlinearities: LMI‐based approach

Due to nonlinear nature of several phase detectors, linear approximation method often leads to performance degradation in many phase‐locked loops (PLLs), particularly when the phase errors are sufficiently large. A third or higher order PLL, in spite of the ability to track a wider variety of inputs and having higher operating‐frequency range, requires more design attention in order to ensure stable tracking. In this work, with the nonlinearities inserted into the system's model, suitable criteria that take into account the nonlinearities' non‐monotonicity, sector and slope bounds are employed to establish robust stability conditions. The result is applicable to any PLLs without order and type restrictions. For Type‐1 PLLs, the resulting condition can be used to search for the maximum stable loop gain, which is also linked to the lock‐in range of the system. In the later part of this work, the focus is devoted towards designing PLLs with high lock‐in range, which is performed via mixing the proposed method with H _∞ synthesis. The searches for the parameters in both PLL analysis and design are expressed in terms of convex linear matrix inequalities, which are computationally tractable. To illustrate the improvement introduced via this approach, several numerical examples and simulations are included with comparisons over conventional methods. Copyright © 2017John Wiley & Sons, Ltd.     

电动汽车复合储能系统的功率分配优化研究

以带有半主动复合储能系统构型的纯电动客车作为研究对象,提出了一种以最小电池耗能及电池功率变化作为目标函数的凸优化方法。在中国哈尔滨城市公交道路工况的基础上,对所提优化方法与基于规则的功率分配策略进行能效及电池功率变化的对比分析。仿真结果表明,在中国哈尔滨城市公交道路工况条件下,采用所提凸优化功率分配策略,电池及超级电容的综合能效分别为93.46%和98.81%,电池功率的均方差为5.3153 k W,较之基于规则的功率分配策略,电池及超级电容的综合能效分别提高了0.74%和0.26%,电池功率均方差降低了46.91%。基于此功率优化分配方法能够有效的改善电动汽车的运行特性。     

高压XLPE电缆击穿的制造因素分析及有效控制

高压电缆尤其是超高压交联聚乙烯(XLPE)电缆在输电运行中一旦被击穿,将对电网造成巨大损害,给国家造成重大损失。同时,企业在生产制造过程中,一旦发生电缆局放值(PC值)超标,甚至耐压测试击穿,给企业造成的经济损失也非常巨大。该文从电击穿理论入手,对高压XLPE电缆的生产制造过程进行了分析研究,找出导致高压击穿及局放值超标的关键要点,如导体毛刺;导体半导电屏蔽带的电导率;模具及流道的磕碰、划伤;材料的洁净和储存;绝缘层中微孔的数量、大小、内应力等。通过对这些容易被忽视,而且又极其重要的生产制造环节进行剖析,并给出控制方法,以达到降低局放值和击穿频率,提高高压XLPE电缆整体品质的目的。     

基于凸规划的机械臂轨迹规划方法

研究了快速求解具有时间约束的机械臂轨迹规划问题,提出了一种基于凸规划的轨迹规划方法。该方法针对机械臂轨迹规划中动力学约束非线性强、时间约束不易处理的问题,首先通过变量替换,将非线性约束转化为线性约束,然后添加新的约束,将原始非凸优化问题转化为凸规划问题,在此基础上,将其写作二阶锥规划（SOCP）形式,使用SeDuMi等优化工具包近似实时求解。该方法具有以下优点：计算高效,凸规划问题能够在多项式时间内得到求解;算法全局稳定,能收敛到全局最优解,不需要提供优化初值;可扩展性强,工业机器人的多种约束以及性能指标如加速度平滑约束、功率等均可扩充。仿真实验表明,与现有方法相比,该方法能够有效提高轨迹规划的效率,机器人的轨迹规划可以近似实时求解。     

基于子空间的对角加载水平估计

Capon波束形成器通常利用对角加载方法来提高稳健性能。然而,对角加载方法的主要缺陷是不容易可靠地获得对角加载水平,从而影响加载效果。由子空间正交理论,噪声与信号子空间相垂直,因此当加载后的导向矢量与真实导向矢量重合时,加载后的导向矢量与噪声子空间垂直。基于这样的特性建立了一个代价函数。分析表明,这个代价函数为一凸问题,通过凸优化软件求解可以很容易地获得合适的加载水平,且与不确定集的参数值无关。仿真结果表明,利用该文获得的加载水平,Capon波束形成器能够有效地提高其稳健性能。     

台面型焊盘的引线键合工艺研究

引线键合工艺是微电子封装的基础工艺,广泛应用于军品和民品芯片的封装。特殊结构焊盘的引线键合失效问题是键合工艺研究的重要方向。文章主要针对台面型焊盘,从热压键合、超声键合、热超声键合原理进行了分析,对台面型焊盘的工艺适应性给出了建议。使用热超声键合工艺进行台面型焊盘的引线键合需要尽可能降低超声功率,避免键合焊盘的机械损伤,并通过平衡其他各变量保证键合点的强度。侧重于热超声键合工艺的应用,分析台面型焊盘与热超声键合过程相关的失效现象,通过样件及小批量试制,对工艺参数进行了优化验证,针对故障件统计分类给出了相应的解决途径。     

GKP外墙外保温薄抹灰贴面砖的探讨

我国南方应用较广的GKP外墙外保温薄抹灰贴面砖出现大面积开裂、脱落问题较多.作者针对这一现象从施工的角度提出解决方法,以供同行参考.     

带式输送机参数化绘图中运输段衔接问题研究

要实现带式输送机的参数化绘图,人们多运用模块化思想,将带式输送机的各组成部件分别成块,选择需要部件进行衔接。作为带式输送机重要部分的运输段、直线段和凸凹弧段的合理衔接成为构成各种工况的关键。通过简单几何原理,将冗繁的衔接简单化,实现带式输送机的参数化绘图。     

Resource allocation for OFDM relay systems with statistical QoS guarantees

By integrating the concept of effective capacity, we propose the resource allocation schemes for subcarrier‐pair based OFDM decode‐and‐forward and amplify‐and‐forward relay systems for quality‐of‐service (QoS) guarantees. The objective is to maximize the system throughput subject to a given statistical delay QoS constraint. First, we pair the subcarriers over the source‐relay channel and the relay‐destination channel by the descending order of the subcarriers' channel gains. Second, by making use of the derived equivalent end‐to‐end channel gains of the subcarrier pairs, we apply joint water‐filling power allocation over the subcarrier pairs and then partition the power of the subcarrier pairs between the source and the relay. We prove that as the equivalent end‐to‐end channel gains of the subcarrier pairs are given, the combination of sorted subcarrier pairing and joint water‐filling power allocation is jointly optimal to maximize the effective capacity. The simulation results show that our proposed schemes achieve the highest effective capacity for OFDM decode‐and‐forward and amplify‐and‐forward relay systems as compared with other existing schemes. The results also verify that our proposed schemes can efficiently provide different levels of delay QoS guarantees, even if under the stringent delay QoS constraints. Copyright © 2012 John Wiley & Sons, Ltd.