The Design Of Adaptive Infrasound Filter Based On LMS Algorithm

As an important aspect of passive detection, acoustic passive localization is one of the focuses of study in modern defense system. The emphasis of this thesis is placed on key technologies of filter in aircraft acoustic localization based on Time Difference of Arrival.According to the theory of the infrasound-signal detection in acoustic localization, the thesis study time and frequency characteristics of the objective infrasound wave, and analysis the attenuation and noise factors in the long-distance communication of the sound wave. Through the simulation,the FIR filter based on LMS adaptive algorithm is chose as the design object.The formula of the step-size is the key factor in the LMS algorithm, some traditional formulas doesn't consider the influence of the input signal and some ones is too complicated in computing. A new non-linear functional relationship between the step-size and the error signal is presented in the thesis, which improved the algorithm performance. The new algorithm considered the influence of the input, and doesn't have operation of exponent, which have better performance.In MATLAB simulation, we can see that the properties of the improved variable step size LMS algorithm such as convergence speed and steady state error are better than those of fixed step-size LMS algorithms. It is shown that algorithm can effectively alleviate fixed step LMS adaptive algorithm in the convergence rate and the contradiction between the steady-state error.Then, the simulation model of LMS algorithm adaptive is built in the SIMULINK, after many experiments, the order of the FIR filter is chose to N = 24, and the step size of LMS algorithm is chose toμ= 0.01. Finally, adaptive filter software design is completed on TMS320F2812 DSP platform, and simulation of software simulator show that the design achieved the desired results.