基于麻雀算法优化GRNN的三维定位算法  

3D positioning algorithm based on GRNN optimized by sparrow search algorithm

作　　者：高媛[1] 阳媛 凌启东[1] GAO Yuan;YANG Yuan;LING Qi-dong(School of Information Engineering,Xuzhou College of Industrial Technology,Xuzhou 221000,China;School of Instrument Science and Engineering,Southeast University,Nanjing 210096,China)

机构地区：[1]徐州工业职业技术学院信息工程学院,江苏徐州221000 [2]东南大学仪器科学与工程学院,江苏南京210096

出　　处：《计算机工程与设计》2022年第11期3149-3158,共10页Computer Engineering and Design

基　　金：国家自然科学青年基金项目(61601123);徐州市科技发展基金项目(KC17132);江苏省青蓝工程人才培养计划基金项目。

摘　　要：为进一步提高室内定位精度,提出一种多种群麻雀算法与广义回归神经网络相结合的室内三维定位算法MSSA-GRNN。MSSA算法种群初始化时采用佳点集及反向学习策略,获得更高质量的初始种群分布;根据适应度值排名,选出优胜子种群和辅助子种群,增强全局搜索能力;在子种群个体更新时,引入渐变加权系数,使算法更快地收敛。经过群内竞争和群间竞争,求得全局最优。利用MSSA优化平滑因子σ的取值,建立最优GRNN神经网络三维定位模型。将仿真结果与其它算法进行比较,证明了所提算法的收敛速度与定位精度均优于其它算法。To further improve the indoor positioning accuracy,an indoor 3D positioning algorithm MSSA-GRNN is proposed,which combines multi group sparrow search algorithm and generalized regression neural network.The MSSA algorithm adopts the good point set and the reverse learning strategy to get the higher quality initial population distribution;According to the fitness value ranking,the superior subpopulation and auxiliary subpopulation are selected to enhance the global search ability;In order to make the algorithm converge faster,the gradual weighting coefficient is introduced when the sub population is updated.The global optimum is obtained through intra group competition and inter group competition.MSSA is used to optimize the smoothing factorσ,the optimal GRNN neural network 3D positioning model is established.Comparing the simulation results with other algorithms,it is proved that the convergence speed and positioning accuracy of the proposed algorithm are better than other algorithms.

关 键 词：无线传感网络 麻雀算法 广义回归神经网络 三维定位 室内定位 多种群 平滑因子 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]