Research On Adaptive Filtering Algorithm For Fusion Of Momentum Terms

The graduation design is a method for adjusting is used to join the new variableleast mean square adaptive filter of the momentum factor adaptive to least mean squareadaptive filter based and introduce the realized based on the performance optimizationof the adaptive filter,which not only has the advantages of fast convergence,steady stateerror and has advantage of small.The autocorrelation function of the adaptive filter cannot be given in advance of signal and noise.adjusting filter coefficient is now timeparameter filter using a moment before it can be obtained automatically,so as to achievethe optimal filtering.This paper deduces the Weiner filter algorithm,and on this basis thefurther introdution of the least mean square algorithm formula and then design anadaptive filter.Through the basic principles of adaptive filter and tell the basicalgorithms of adaptive filter.,puts forward the theory basis to improve the momentumand adaptive filter algorithm,the basic design thought and algorithm are also variablefactor adaptive filtering algorithm momentum,is simulated using the Matlab assemblersoftware,based on the Matlab diagram analysis and different algorithms and comparingthe improved algoriyhm,obtained in various indicators.The simulation result is ideal,infact,improved adaptive filtering algorithm has the convergence rate and the steady-stateerror on the edge,to verify the effectiveness of the algoritnm.Through Matlabsimulation,to work principle to help learners grasp the adaptive filter, to understand itsbasis structure,easy to master the adaptive filtering techniques make it easier.This paperis divised to three typical—Weiner filtering algorithm in adaptive filter,the steepestdescent method,LMS algotithm,LMS algorithm was proposed to track is more suitablefor the external changing environment and quickly adjust the filter coefficients. In thispaper, the traditional LMS algorithm for the simulation of advance, gives the simulationresults of Matlab. Then on the basis of adding the momentum item weight vectorupdating, the momentum LMS algorithm improved optimization concept, adding momentum factor, by optimizing the momentum factor, with better convergenceproperties of the algorithm.