基于反正切函数的LMS自适应滤波算法及应用  

LMS Adaptive Filtering Algorithm Based on Arctangent Function and its Application

作　　者：张晓亚 刘建勋 倪元相 吴亮[1] 彭金奇[1] ZHANG Xiaoya;LIU Jianxun;NI Yuanxiang;WU Liang;PENG Jinqi(College of Mechanical and Electrical Engineering,Guangzhou 510510,China;School of Mechanical Engineering,Southeast University,Nanjing 211189,China;Science and Technology College of Guangdong Polytechnic,Zhaoqing 526200,Guangdong,China)

机构地区：[1]广东工贸职业技术学院机电学院,广州510510 [2]东南大学机械工程学院,南京211189 [3]广东理工学院科技处,广东肇庆526200

出　　处：《弹箭与制导学报》2022年第1期114-117,共4页Journal of Projectiles,Rockets,Missiles and Guidance

基　　金：广东省普通高校工程技术中心项目(2021GCZX014)资助。

摘　　要：在对无线电引信进行信号处理的过程中,现有的最小均方差(LMS)算法不能有效的实现较好的收敛性能和稳态特性。为了改善LMS滤波算法的不足,综合考虑迭代步长与误差的关系,前一次迭代步长与当前迭代步长的关系以及当前迭代步长的变化情况,提出了一种新的基于迭代步长、误差和动态约束迭代步长的数学模型。通过仿真,分析新算法中的关键参数对滤波性能的影响,并确定其最优的参数。在信号的滤波处理中,从仿真效果可以看出:新算法具有很好的收敛性能和稳态特性;与原有算法相比,抗干扰能力提高3.5倍。In the process of signal processing for radio fuzes,the existing least mean square(LMS)algorithm cannot effectively achieve better convergence performance and steady-state characteristics.In order to improve the deficiencies of the LMS filtering algorithm,comprehensively considering the relationship between the iteration step size and the error,the relationship between the previous iteration step size and the current iteration step size,and the change of the current iteration step size,a new iteration based on the iteration step size is proposed.Mathematical model of error and dynamic constraint iteration step length.Through simulation,the impact of the key parameters of the new algorithm on the filtering performance is analyzed in detail,and the optimal parameters are determined.In the signal filtering process,it can be clearly seen from the simulation effect that the new algorithm has good convergence performance and steady-state characteristics.Compared with the original algorithm in terms of anti-interference,the anti-interference ability is increased by 3.5 times.

关 键 词：LMS 滤波算法 收敛性能 稳态特性 

分 类 号：TN911.72[电子电信—通信与信息系统] TP391.9[电子电信—信息与通信工程]