无线传感器网络中传感器节点定位的递归加权最小二乘优化方法

采用最小二乘优化算法,研究了基于接收信号强度测量的信号源定位问题.首先,建立了路径损耗的线性化模型;然后,介绍了加权最小二乘法,并提出了无线传感器网络中传感器节点定位的递归加权最小二乘优化方法.最后将本文提出的迭代加权最小二乘法与基于接收信号强度（RSS）的2种算法进行比较,仿真结果表明提出的算法在定位精度方面有更好的表现.      

李雅普诺夫方程的新解法

利用对数欧氏度量的方法给出李雅普诺夫方程的新解法.介绍了李雅普诺夫方程的由来,介绍对称正定矩阵流形的黎曼度量以及对数欧氏度量下的距离函数,给出求解李雅普诺夫方程的迭代公式,并给出模拟仿真的结果.      

Drag reduction through self-similar bending of a flexible body

The classical theory of high-speed flow predicts that a moving rigid object experiences a drag proportional to the square of its speed. However, this reasoning does not apply if the object in the flow is flexible, because its shape then becomes a function of its speed--for example, the rolling up of broad tree leaves in a stiff wind. The reconfiguration of bodies by fluid forces is common in nature, and can result in a substantial drag reduction that is beneficial for many organisms. Experimental studies of such flow-structure interactions generally lack a theoretical interpretation that unifies the body and flow mechanics. Here we use a flexible fibre immersed in a flowing soap film to measure the drag reduction that arises from bending of the fibre by the flow. Using a model that couples hydrodynamics to bending, we predict a reduced drag growth compared to the classical theory. The fibre undergoes a bending transition, producing shapes that are self-similar; for such configurations, the drag scales with the length of self-similarity, rather than the fibre profile width. These predictions are supported by our experimental data.     

曲指数族流形在无线传感器网络定位中的应用

利用信息几何中的统计流形理论和自然梯度流形学习定位方法,研究了基于接收信号强度（RSS）的无线传感器网络自定位问题.首先,通过概率密度函数构造了一个曲指数族定位模型;然后,针对给定初始状态值的未知目标节点定位问题,结合梯度下降法,提出了基于此模型的最优非线性估计方法及其改进算法.梯度下降法的良好性质和仿真结果表明,这些算法有很好的收敛效果和更高的定位精度.