��������Ӧ��ζ�����˲�

????????????д?????????????????????????ζ?????????(UPF)?????????????????????????μ???????????ζ?????????(AR??UPF)???÷??????÷???????????????????????????????UKF??????????????????????Э????????????????????????????????·????????Ч??????????????????????????????     

�����Ƚ��ֻ�������Bayes���Ʒ��ڵؿ��α�����е�Ӧ��

??????????????????α?????GPS???ι???????????α?????????????????????????????????????????????????????????????????????????в??????????????????????????????????????????α???????????????????ι????????????α???????????????????Ч?????????????????????????????????Bayes?????????????????α?????????????????????????????????????????????????????????????????????????????????????????????Щ???????????????????????????????????????????????????GPS????????м??????????????????????????Ч???????????????????????????????????????     

Effective force testing using a robust loop shaping controller

Effective force testing (EFT) is one of the force‐based experimental methods used for performance evaluation of structures that incorporate dynamic force control using hydraulic actuators. Although previous studies have shown successful implementations of force control, controllable frequency ranges are limited to low frequencies (10 Hz). This study presents the EFT method using a robust loop shaping force feedback controller that can extend the frequency range up to 25 Hz or even higher. Unlike the conventional PID controllers, loop shaping controllers can provide robustness for a high level of force measurement noise. This study investigates the dynamic properties of hydraulic actuators and the design of a loop shaping controller that compensates for control–structure interaction and suppresses the effect of oil‐column resonance. The designed loop shaping controller was successfully implemented into an EFT setup at the Johns Hopkins University. An experimental investigation of the loop shaping controller was performed under step, random, and earthquake force loadings. Experimental results showed that the loop shaping controller provided excellent force tracking performance and robustness for dynamic force loadings. It was also shown that the loop shaping controller had the gain margin of 9.54 dB at the frequency of 28 Hz. Copyright © 2012 John Wiley & Sons, Ltd.     

Multidegrees‐of‐freedom effective force testing: a feasibility study and robust stability assessment

This paper presents a feasibility study of multidegrees‐of‐freedom effective force testing (MDOF‐EFT). The study is intended to facilitate the development of a force feedback controller and investigation of performance as well as robustness of MDOF‐EFT. First, the dynamics of MDOF‐EFT systems are analytically investigated. Analytical transfer functions of the control plant, the valve‐to‐force relations, showed that the plant is dynamically coupled and the natural frequencies of test structures are the transmission zeros of the plant. Using a set of model parameters from a previous study, a case study that includes controller design, numerical simulations and robust stability assessment is performed. A decoupling loop shaping (DLS) controller consisting of a pseudo inverse of the plant and second‐order loop shaping controllers is adopted as the force feedback controller. It is shown that the DLS controller provides a stable control system while successfully decoupling the control loops and compensating the control‐structure interaction. Numerical simulations demonstrate that the DLS controller enables tracking of static and dynamic forces for multiple actuators. Robust stability of MDOF‐EFT with the DLS controller is assessed using Monte Carlo simulation. The stochastic simulation results show that the DLS controller is stable and robust, providing sufficient stability margins for uncertain models with maximum 50% errors in the estimated system parameters. This paper demonstrates that MDOF‐EFT is feasible with the DLS controller and can be implemented in experimental laboratories. Copyright © 2013 John Wiley & Sons, Ltd.     

朱集矿首采面观测站高程连接测量与数据处理

结合淮南朱集煤矿首采面地表移动观测站的高程连接测量第一次全面观测,设计了首采面地表移动的三等水准控制网,利用抗差估计模型剔除粗差,结合最小二乘估计模型对观测结果进行了处理,并从精度、可靠性和置信度等三个方面进行分析并得出结论。为以后采集各个阶段的地表移动数据,求出参数并且得出结论提供有效的基础数据,从而为朱集矿的后期采煤提供有效的技术保证。     

基于抗差估计的GPS／CoMPASS组合实时差分定位算法

在GPS／cOMPASS组合差分定位基础上,引入抗差估计方法,通过降权调整合有粗差观测值对定位的影响,利用路测试验,对比RTK的厘米级定位,结果表明：抗差GPS／COMPASS组合实时差分定位优于非抗差定位,平面精度在2．0m左右。     

丹麦法稳健估计在隧道椭圆拟合中的应用

隧道断面在设计时为一标准圆形,受到施工及外部环境的影响变形为偏心率很小的椭圆。减小粗差影响、精确确定椭圆的几何参数是其研究的重点。通过对上海某一隧道断面的测量试验,得出用丹麦法稳健估计可以有效减小粗差的影响,而且拟合值可以很好地逼近无粗差情况下的最小二乘拟合值。采用Matlab 语言实现算法,并详细介绍了用二项式变换求解椭圆参数的方法。     

GPS/Doppler导航随机模型的移动窗口实时估计算法

引进时间遗忘因子和观测冗余度因子,有效地平衡移动窗口内不同时刻的观测数据及其冗余情况对单位权方差估值的贡献,改进单位权方差的移动开窗实时估计算法.采用载噪比模型确定观测权阵,等价权抗差估计方法处理粗差.实测车载GPS/Doppler数据的处理结果表明:采用本文算法显著提高GPS/Doppler的导航精度与可靠性.     

超高阶次勒让德函数递推计算中的压缩因子和Horner求和技术

摘要：超高阶次（如2000阶次）缔合勒让德函数值的递推计算,在接近两极时达到极大的数量级（超过10的数千次方）,这导致现有递推方法在计算缔合勒让德函数值及其导数值时失效。我们通过插入压缩因子技术,提出一个修改的递推算法,并结合使用Horner求和技术计算球谐级数的部分和。试验表明,该算法至少可以计算到3600完全阶次的球谐级数式,优于现有结果。     

抗差自适应卡尔曼滤波在GPS精密单点定位中的应用

卡尔曼滤波是GPS精密单点定位中最常用的参数估计方法.理论上讲,随着观测数据的增多,通过卡尔曼滤波可以得到更精确的状态估值,但有时由于状态异常或发生大的周跳等原因,使得状态估值与实际状态之间的误差较大,滤波会发生发散现象,为此,本文提出了将抗差自适应滤波模型运用到精密单点定位中,通过算例分析显示,该方法将显著提高定位结...