WIFI位置指纹定位作为目前常见的室内定位方法,存在接收信号强度(received signal strength,RSS)波动和时变等问题,导致定位精度不高。文章为此设计了一种采用结合卡尔曼滤波的方差修正加权K最近邻(weighted K-nearest neighbor,WKNN)...WIFI位置指纹定位作为目前常见的室内定位方法,存在接收信号强度(received signal strength,RSS)波动和时变等问题,导致定位精度不高。文章为此设计了一种采用结合卡尔曼滤波的方差修正加权K最近邻(weighted K-nearest neighbor,WKNN)算法的室内定位方法。离线阶段,经过卡尔曼滤波后,选择数据的方差和均值作为反映RSS变化的特征值;在线阶段,通过采集的信号均值计算近似方差,对欧式距离进行权重修正,最后选择K个最近邻点确定待定点位置。实验结果表明:该文采用的定位方法平均定位精度达到1.248 m,相比于传统的WKNN室内定位方法,平均定位精度提升了20.3%;对比K-均值聚类结合动态加权K最近邻算法(K-means-EWKNN),平均定位精度提升了8.9%。展开更多
The effect of signal modulating noise in bistable stochastic resonance systems was studied theoretically and experimentally. A mathematical analysis was made on the bistable stochastic resonance model with small syste...The effect of signal modulating noise in bistable stochastic resonance systems was studied theoretically and experimentally. A mathematical analysis was made on the bistable stochastic resonance model with small system parameters. An analogue circuit was designed to perform the effect. The effect of signal modulating noise was shown in the analog simulation experiment. The analog experiment was conducted for two sinusoidal signals with different frequencies. The results show that there are a sinusoidal component corresponding to the input sinusoidal signal and a noise component presented as a Wiener process corresponding to the input white noise in the system output. By properly selecting system parameters, the effect of signal modulating noise can be manifested in the system output.展开更多
基金Project (10276032) supportedjointly by the National Natural Science Foundation of China and by the Science Foundationof China Academy of Engineering Physics NSAFproject(2005038228) supported by Postdoctoral Science Foundation of China projectsupported by the Postdoctoral Science Foundation of Central South University(2005)
文摘The effect of signal modulating noise in bistable stochastic resonance systems was studied theoretically and experimentally. A mathematical analysis was made on the bistable stochastic resonance model with small system parameters. An analogue circuit was designed to perform the effect. The effect of signal modulating noise was shown in the analog simulation experiment. The analog experiment was conducted for two sinusoidal signals with different frequencies. The results show that there are a sinusoidal component corresponding to the input sinusoidal signal and a noise component presented as a Wiener process corresponding to the input white noise in the system output. By properly selecting system parameters, the effect of signal modulating noise can be manifested in the system output.
