基于自然选择粒子群的时钟同步算法

发布时间：2018-08-31 13:10

【摘要】：为了提高无线传感器网络的时钟同步精度,避免因不可靠数据和网络拓扑结构变化在同步过程中产生的同步误差,建立了基于自然选择粒子群融合算法仿真模型.首先,采用卡尔曼滤波算法,将多个传感器的测量数据进行局部滤波处理,去除测量数据中最底层的冗余信息,提高测量数据的准确性;其次,采用自然选择粒子群算法建立数据融合模型,计算网络的最优融合估计;最后,对自然选择粒子群融合算法模型进行仿真,结果表明:该算法的融合模型不仅能够对网络中的5个节点时钟进行有效融合,而且可以提高节点晶振的时偏和频偏的融合精度,形成一个具有较高稳定度的虚拟时钟,其时偏和频偏的融合精度主要集中范围分别为10-5和10-7,与传统自适应融合算法相比较,它的时偏和频偏融合精度提高了1个数量级.
[Abstract]:In order to improve the clock synchronization accuracy of wireless sensor networks and avoid the synchronization errors caused by unreliable data and network topology changes, a simulation model based on natural selection particle swarm optimization algorithm is established. Firstly, the Kalman filter algorithm is used to process the measurement data of multiple sensors, which removes the redundant information at the bottom of the measurement data, and improves the accuracy of the measurement data. The natural selection particle swarm optimization algorithm is used to establish the data fusion model, and the optimal fusion estimation of the network is calculated. Finally, the natural selection particle swarm optimization algorithm model is simulated. The results show that the fusion model of the algorithm can not only effectively fuse the five node clocks in the network, but also improve the fusion accuracy of the time offset and frequency offset of the node crystal oscillator, and form a virtual clock with high stability. The fusion accuracy of time-offset and frequency-offset is 10-5 and 10-7, respectively. Compared with the traditional adaptive fusion algorithm, the fusion accuracy of time-offset and frequency-offset is improved by one order of magnitude.
【作者单位】：哈尔滨工程大学信息与通信工程学院;
【基金】：国家自然科学基金资助项目(NSFC-61571146)
【分类号】：TN929.5;TP18;TP212.9

【相似文献】