The Control System Design And Implementation Of Adaptive Active Noise Cancelling Headset Based On DSP

At present, almost all of active headset controller is based on analog circuit. The transfer function parameters of the controller are fixed and cannot be adjusted, and the actual noise is non stationary and time varying characteristics, and some uncontrollable parameters in transfer function or noise reduction space often change with time. In addition, it is difficult to achieve multi-channel systems dealing with complex noise. Therefore, this thesis designed an adaptive active headset control system based on DSP.First of all, adaptive active headset based on DSP in the market positioning and prospect was analyzed, and clear the applications, and analyzed the primary noise and noise reduction space characteristics in the target situation, formulated the system control strategy and control tasks. Then, from the performance point of view, the control method, the controller structure and the algorithm of the system are analyzed, and the solution of the control system is determined.Secondly, the model of feed-forward single channel system was analyzed. On the basis of this, the Fx LMS algorithm is derived, the performance of the algorithm is analyzed, and the modeling problem of the secondary path was discussed. Then, the FxLMS algorithm was programmed based on the MATLAB software, and a feed forward single channel system model was built in the Simulink. Simulation results show that adaptive active controller based on the transverse structure of FIR filter and FxLMS algorithm in handling narrowband primary noise and feed forward single channel system can gain better noise reduction effect. In addition, the higher filter orders and convergence factor makes the convergence system faster, but the stability of the system affected.Finally, the control system of adaptive active noise cancelling headset was designed and realized based on DSP, and built the experimental platform. Experimental results show that the system can automatically track the statistical characteristics of the noise, and be real-time to adapt to changes in the noise environment, and achieved obvious noise reduction effect, and the feed-forward structure ensures better stability.