基于FastICA的无人机声学检测方法

UAV acoustic detection method based on FastICA

随着无人机(UAV)的广泛应用,无人机“黑飞”等问题也随之而来。针对强干扰环境下的无人机声音识别问题,提出一种基于FastICA算法的无人机识别方法。使用FastICA算法提取出无人机声音,再对无人机声音进行识别,从而提高了无人机声音检测方法的抗干扰能力。实验结果表明:所述方法在多种声源混合的情况下,仍能较好地识别无人机声音,并对不同型号的无人机均有较好地识别效果。同时考虑了识别距离对识别率的影响,结果表明:随着识别距离变大,所述算法仍能较好地识别无人机。

With the wide application of unmanned aerial vehicle(UAV),problems such as unlicensed UAV navi-gation also follow.Aiming at the problem of UAV acoustic detection in strong interference environment,an UAV acoustic detection method based on FastICA is proposed.FastICA algorithm is used to extract the sound of UAV,and detect UAV audio,so as to improve the anti-interference ability of UAV sound detection method.Experimental results show that the method can still achieve UAV acoustic detection well even if in the case of multiple sound sources,and have a good recognition effect for different types of UAV.At the same time,the influence of recognition distance on recognition rate is considered.The results show the proposed algorithm can still identify the UAV well with the increase of distance.