Test Platform For Active Noise Control System In Enclosure

In recent years, vehicle consumers seek for quietness inside the car. While low fre-quency vehicle interior noise could hardly be reduced using conventional passive methodunder vehicle lightweight trend. Active noise control technique, hence, helps to solvethis problem, achieving good balance between vehicle economy and comfort, due to lackof bulky and expensive absorption materials. Moreover, the vehicle technologies are de-veloping rapidly. The research and development on downsizing, cylinder-on-demand,hybrid and electrical vehicles bring about new challenges to vehicle NVH performance.Active noise control, again, can ofer support to these new technologies. Therefore, it isof practical importance to study on active noise control technique.In order to study systematically on the vehicle ANC technique, as well as to meetthe demand of validating the ANC control efect under diferent configurations, a studyon ANC test platform is conducted. According to the existed interior noise reproductionsystem, an ANC system modeling acoustic characteristics of vehicle is built up. Afteranalyzing the demand for the system, a set of hardware is constructed; at the mean time,the corresponding software is developed.Considering delay will cause divergence and instability to the system, a modifiedadaptive control algorithm with delay compensation is proposed based on FXLMS. Ef-forts are further made on the delayed secondary path estimation as well as on the adaptivefiltering coefcient update under block data acquisition. Moreover, the step size in thecontrol algorithm is chosen according to simulation and the influence of sample delay,which in a certain range will not deteriorate the noise suppression, is analyzed. Tak-ing advantage of this feature, a semi-online delay recognition-compensation method isproposed.A single channel ANC system is run under the test platform built. Maximum26dBat control sensor position is achieved in suppressing single frequency primary field, whichvalidates both the modified adaptive control algorithm with delay compensation and thesemi-online delay recognition-compensation method.Considering the low computing efciency of convolution, which may afect the real-time control in a multi channel system. An improvement of computing efciency with1/50of the original computing time is achieved by using fast Fourier transform. Then, 10～16dB is achieved at each control sensor position by applying multi channel ANCsystem which uses vehicle loudspeakers as secondary sources. In addition, other than thecontrol points, sound suppression is also obvious in other points in the enclosure, with anoverall suppression of5dB. At last, an ANC test for time-variant vehicle interior noise isconducted, with suppression of5～15dB at control points and3～5dB overall.