针对群组移动节点定位算法普遍基于不切实际的假设,存在普适性欠佳和精度不高的问题,提出一种基于运动参数预测的群组移动节点定位算法。该算法根据群组移动节点具有相似运动的特点,运用Hermite插值多项式预测、过滤节点运动参数。为确保定位精度,应对节点移动性带来的采样区域变化,运用预测节点运动参数构建粒子有效采样区域;为节省时间开销,基于采样粒子真实分布与其极大似然估计值之间的最大K-L(Kullback-Leibler)距离确定能够满足不同采样区域的最少粒子数目;为改善算法收敛性,运用预测运动参数创建滤波公式,并选取优质粒子参与节点位置估计。在与经典算法MCL(Monte Carlo localization)法和加权最小二乘法的MATLAB对比实验中,分析了节点移动速度、自由度、K-L距离阈值、采样方格边长对定位精度的影响。结果表明,较上述算法,本算法的定位误差和时间开销较小,无须锚节点辅助,普适性较好。
Pulse decomposition has been proven to be efficient to analyze complicated signals and it is introduced into the photo-acoustic and thermo-acoustic tomography to eliminate reconstruction distortions caused by negative lobes.During image reconstruction,negative lobes bring errors in the estimation of acoustic pulse amplitude,which is closely related to the distribution of absorption coefficient.The negative lobe error degrades imaging quality seriously in limited-view conditions because it cannot be offset so well as in full-view conditions.Therefore,a pulse decomposition formula is provided with detailed deduction to eliminate the negative lobe error and is incorporated into the popular delay-and-sum method for better reconstructing the image without additional complicated computation.Numerical experiments show that the pulse decomposition improves the image quality obviously in the limited-view conditions,such as separating adjacent absorbers,discovering a small absorber despite disturbance from a big absorber nearby,etc.
