The Research And Implementation Of Automobile Wind Noise ANC System Based On FxLMS Algorithm

With the rapid development of the transportation industry and economy,people's requirements for the ride comfort of cars are constantly increasing.The wind noise in the driving of the car is inevitable,and can reach 90 dB in serious cases,which not only interferes with the communication between the passengers but also causes hearing damage to the passengers.Most methods for solving wind noise are passive materials for wind noise,such as acoustical glass or thicker windshields,but this noise reduction method only works well for high frequency noise and is lighter than cars.The goal of quantification is contrary.Therefore,according to the principle of acoustic superposition,it is better to use the active noise control system to generate artificial anti-wind noise signals to cancel the wind noise inside the vehicle.Firstly,this paper analyzes the composition and transmission path of wind noise in the car,introduces the basic principle of ANC system,and discusses the structural characteristics of hybrid,feedforward and feedback ANC systems.Based on the research of ANC system adaptive algorithm,the theoretical derivation of LMS and FxLMS algorithms is carried out,and the performance indexes such as convergence,steady state and complexity of the algorithm are analyzed in depth.The FxLMS adaptive algorithm is applied to single-channel feedforward active.The ANC system has good applicability.Secondly,for the ANC system,there is a secondary channel identification problem in practical applications.Through theoretical derivation and analysis of existing secondary channel modeling methods,it is proposed to use the main channel for the secondary channel path in the vehicle wind noise active noise control system.Secondary channel optimal variable step size algorithm.The algorithm uses white noise as the training signal of the secondary channel modeling filter,and uses the LMS algorithm to adapt the real secondary channel path to obtain the estimated secondary channel path transfer function.The improved FxLMS algorithm achieves better results in secondary channel modeling accuracy,ANC system steady-state performance and convergence speed by Matlab simulation.Finally,based on the improvement of FxLMS algorithm based on automotive wind noise ANC system,the hardware implementation platform of ANC system is established,and FPGA(field programmable gate device)is used as ANC system controller to improve the hardware circuit structure of the improved FxLMS algorithm.verification.A system architecture for FPGA fast implementation is designed,and the structure is simulated by Verilog HDL hardware description language.The simulation results show that the improved FxLMS algorithm can quickly and accurately model the secondary channel path.This allows the ANC system to counteract the effects of wind noise.