在噪声抵消应用中自适应滤波算法性能仿真比较.pdfVIP

22 4 Vol1 22 No14 2008 8 JOURNAL OF ELECTRON ICMEA SUREMENT AND INSTR UMENT # 57 # 1 1 2 罗俊海 叶丹霞 李录明 ( 11 , 6 10054; 21 , 610059) : RBF , BP RBF , , , BP RBF , , , : , , , + : TP273 14 : A : Simulation Comparison ofAdaptive FilteringAlgorithms in Noise Cancellation Application 1 1 2 Luo Junhai Y Danx ia Li Lum ing ( 11Coll g of Com put r, Un iv rsity of E l ctron ic Sci nc and T chnology of Ch ina, Ch ngdu 6 10054, China; 21Coll g of In form ation Engin ring, Ch ngdu Un iv rsity of T chnology, Ch ngdu 6 10059, Ch ina) Abstract: Th th ory of nois canc ling and th m thod for xtracting d sir d signal from strong background nois using adaptiv filt ring ar d scrib d. Th configuration of fuzzy n utral n tw ork is also pr s nt d by comb-i n ing fuzzy logic and n utral n twork bas d on th ir qu ival nc . And th n BP n uraln twork algorithm, RBF n u- ral n twork algorithm and fuzzy n ural n tw ork algorithm ar compar d. Comput r smi ulation r sults show that all of th s adaptiv algorithm s can mi prov th d t ction ofw ak signal in strong background nois . In comparison, th p rformanc of fuzzy n ural n wt ork algorithm is much b tt r than thos of BP n ural n wt ork algor ithm and RBF n ural n tw ork algorithm. B sid s much fast r conv rg nc sp d, th b havior of fuzzy n ural n wt ork filt ring co- ffici nts is much mor stabl; and fuzzy n ural n wt ork has fast r in itial conv rg nc rat , low r maladjustm nt nois and b tt r robustn ss against nois