Adaptive noise cancellation using soft computing approach
Zheng Rong Li · 2006
5.17 (a) Online reproduction error of the last 200 samples and (b) online recovered signal of the last 200 samples 167 5.18 Procedure of weights update of the first fuzzy rule in the rule base. 167 5.19 Power spectral density, (a) Information signal (handel.m) s(k), (b) noise source (train.m) x(k), (c) additive noise (through the changing nonlinear channel) n(k), and (d) distorted signal d(k) 168 5.20 Power spectral density of information signal s(k) (solid line) and online recovered signal e(k) (dotted line) 168 5.21 Model validity test. The dotted lines correspond to the 95% confidence bands, (a) V Te (t), (b)V x 2 e (t), and (c)*3es(0 169 6.1 Plant identification in the series-parallel mode 178 6.2 Multiple-indepcndent-adapting scheme using the series-parallel mode. 181 6.3 Training results of Example l.(a) Growth of EBF neurons, (b) desired output, (c) online identified output, and (d) online training error 185 6.4 Testing result of the trained fuzzy system for Example 1 185 6.5 Training results of Example 2. (a) The generation of fuzzy rules and (b) online training error 187 6.6 Testing result of Example 2 188