光纤陀螺随机误差特性ThêoH辨识

随机误差特性分析是光纤陀螺进入工程应用阶段的关键技术.针对Allan方差、总方差、Thêo1 (Theoretical variance #1)等分析方法的不足,以Thêo1为基础,通过偏差消除获得ThêoBR(Bias-removed ver?sion of Thêo1),并将其与Allan方差合成,得到光纤陀螺随机误差特性分析的ThêoH(Hybrid-ThêoBR)法.通过计算重叠Allan方差、总方差、Thêo1以及ThêoH的等效自由度,结合仿真和实测信号的随机误差特性分析实验,证明ThêoH法不仅提高了Allan方差估计的置信度,且解决了Thêo1估计有偏的问题,表明了该方法是目前能够有效辨识光纤陀螺随机误差特性的最佳方法.     

提升容器横向偏移惯性测量法的误差分析与校正

针对超深矿井提升装备需要增加提升容器横向偏移测量的问题,根据其小量程、高精度、抗干扰的测量要求,提出一种基于MEMS传感器的惯性测量方法.采用惯性测距原理进行横向偏移测量,分析了倾斜角、随机误差和速度残值三种主要误差源,提出了对应的动态均值、阈值设置和速度残值校正与补偿三种校正方法,结合提升容器的运行特点消除累积误差,并在液压移动平台上进行实验.结果表明:误差校正方法可有效抑制误差,横移测量精度在5 mm以内,惯性测量方法是一种超深矿井提升容器横向偏移测量的有效方法.     

一种基于EEMD的新型谐波误差分析策略

针对电力系统中严重的谐波干扰问题,为提高其运行的安全有效性,去除谐波干扰并提高系统稳定性能成为研究热点。电网中常存在频率强烈波动或者白噪声强烈干扰的现象,虽说三谱线插值的FFT算法(基于Kaiser窗)的抗干扰能力很强,谐波参数估计也可以有较高的精准度,但是远不及使其与平均经验模态的总体水平来分解谐波的检测算法(EEMD)与直线度误差分析相结合。应用直线度误差来分析信号的谐波成分,有机结合了三谱线插值的FFT(基于Kaiser窗的)谐波检测分析法。通过对整体谐波分量的在线监测,不仅能够提高稳态信号的检测精度,还能提高系统的动态性能。本分析策略大限度地提高了计算精度以及系统稳定性。通过仿真分析得知,所述方法可以使系统具有较强的抑制频谱泄露能力,并能够显著降低栅栏效应的发生概率,进而使电力谐波信号的幅值误差、初相位误差和频率误差分别低于0.004 1%、0.002 4%和0.003 9%。     

Robust Adaptive Neural Control for a Class of Time‐Varying Delay Systems with Backlash‐like Hysteresis Input

This paper proposes a robust adaptive dynamic surface control (DSC) scheme for a class of time‐varying delay systems with backlash‐like hysteresis input. The main features of the proposed DSC method are that 1) by using a transformation function, the prescribed transient performance of the tracking error can be guaranteed; 2) by estimating the norm of the unknown weighted vector of the neural network, the computational burden can be greatly reduced; 3) by using the DSC method, the explosion of complexity problem is eliminated. It is proved that the proposed scheme guarantees all the closed‐loop signals being uniformly ultimately bounded. The simulation results show the validity of the proposed control scheme.     

A new robust observer-based adaptive type-2 fuzzy control for a class of nonlinear systems

In this paper, a novel robust observer-based adaptive controller is presented using a proposed simplified type-2 fuzzy neural network (ST2FNN) and a new three dimensional type-2 membership function is presented. Proposed controller can be applied to the control of high-order nonlinear systems and adaptation of the consequent parameters and stability analysis are carried out using Lyapunov theorem. Moreover, a new adaptive compensator is presented to eliminate the effect of the external disturbance, unknown nonlinear functions approximation errors and sate estimation errors. In the proposed scheme, using the Lyapunov and Barbalat's theorem it is shown that the system is stable and the tracking error of the system converges to zero asymptotically. The proposed method is simulated on a flexible joint robot, two-link robot manipulator and inverted double pendulums system. Simulation results confirm that in contrast to other robust techniques, our proposed method is simple, give better performance in the presence of noise, external disturbance and uncertainties, and has less computational cost.     

Improved phase performance for rotman lens

Rotman lenses are used to obtain multiple beams from a single array. Although the beams produced by the feed antennas at focal points have no path length errors, the beams produced by feed antennas at off focal points may have large path length errors. These path length errors cause deterioration in the multiple beams. In this article, two novel methods are introduced to obtain feed curves which reduce the path length errors of off focal feed points significantly, compared with the commonly used circular and elliptical feed curves. The first method obtains feed curve points based on having zero path length error at three chosen points of the radiating array for each beam direction. The second method uses the particle swarm optimization method for obtaining optimum feed points for each beam direction. The results show that there is a very significant drop in the level of the maximum path length errors (in the order of about 1:4). © 2012 Wiley Periodicals, Inc. Int J RF and Microwave CAE 23: 634–638, 2013.     

六轴工业机器人的参数辨识方法

机器人末端位姿的精度依赖于各关节几何参数的精度。为了提高机器人的位姿精度,提出了一种简单实用的机器人参数辨识方法。利用D-H 算法建立的机器人运动学方程,得到了末端位姿与各关节参数的关系。为了解决辨识过程中的复杂计算问题,设计了一种采用拆分参数的迭代方法和相应的实验方法,运用激光跟踪仪对位姿进行精确测量,求得实际几何结构参数。实验结果验证了所提出方法的有效性。     

一种面向多核处理器I/O系统软错误容错方法

集成电路制造工艺的飞速发展,使得集成电路的特征尺寸不断减少和集成度不断提高,造成集成电路对工作环境的影响越来越敏感,发生软错误的几率不断增加,对可靠性造成重要影响。随着微处理器进入了多核时代,丰富的片上资源给软错误加固带来了很好的机遇。本文针对多核处理器中I/O系统软错误,提出了一种基于多核处理器的软件Scrub方法对软错误进行加固。测试结果表明,我们提出的软错误容错方法可以大大提高I/O系统的可靠性。