Study On Convex Combination Of Filtered-X Affine Projection Algorithm In Active Noise Control System With Nonlinear Secondary Paths

With the rapid development of modern society,noise pollution problem is increasingly more prominent,therefore,noise control is an urgent need for work.Traditional passive noise control technology has good control effect on high frequency noise,but not to the inhibitory effect on low frequency noise,and most of the human living environment is in the low frequency range,thus active noise control technology is born at the right moment,and with automatic tracking and the statistical properties of the primary noise controller features for time-varying adaptive active noise control system is a hot research topic again.In adaptive active noise control system,although linearity models simplify the mathematic problem and often permits a detailed system analyses,it cannot evaluate the performance of the system practically.Signal converters(A/D and D/A),power amplifiers,and transducers(speakers or actuators)can nonlinearly transform digital electrical signals into analog electrical or mechanical signals.Small nonlinearity is shown to significantly affect algorithm performance.In addition,the impact on the performance of adaptive active noise control filter is influenced by in addition to the adaptive control algorithm,and its structure.Thus,the main contributions of this thesis are summarized as follows:Firstly,establishing the single channel feedforward adaptive active noise control system model,and analyses the core of this system,i.e.the adaptive filter and its corresponding algorithms,including made a brief introduction and comparison of the simulation performance for filtered-X least mean square algorithm(FXLMS algorithm),filtered-X normalized least mean square algorithm(FXNLMS algorithm)and filtered-X affine projection algorithm(FXAP algorithm).Secondly,being focus on FXAP algorithm considering affine projection algorithm(AP algorithm)can make a better balance in convergence speed and maladjustment,and bringing into saturation nonlinear functiong g(·)at the secondary path,further studying the steady state and transient performance and computational requirements of FXAP algorithm of nonlinear secondary channel in single channel active noise control system within the scope of real Numbers from the perspective of energy conservation,and the simulation results verify the theoretical analysis.Finally,in order to improve the performance of the filter,FXAP algorithm based on convex combination in the system of single channel active noise control system with nonlinear secondary path is proposed,and the computing requirement is briefly introduced and the simulation analysis has been done for the performance in the non-stationary state and Bernoulli-Gaussian pulse noise environment.