基于SRP和GA的声源定位方法

Sound Localization Method Using SRP and GA Algorithm

目前在枪声的定位系统中,大多采用到达时间差(DTOA)的方法对声源方向进行估计,这种方法没有用到阵列整体的优势,缺乏稳健性。而基于可控响应功率(SRP)的波束形成技术,具有较强的稳健性、抗噪性,以及具有一定的空域滤波性能,能够提高信号的信噪比,增加探测距离。在确定声源位置时,一般采用最优化算法,最小二乘法容易陷入局部极值点,而基于自然选择法则的遗传算法(GA)模拟生物进化过程,具有并行计算的特点,是一种全局搜索算法,能够得到全局最优解。将基于可控响应功率的波束形成技术和遗传法算法应用于枪声定位系统中,并通过实际的试验数据对此方法进行了验证,从结果来看具有较高的精度。

Currently, gunshot acoustics directional system relys on a simple calculation of differences of time-of-arrival to estimate angle-of-arrival. This approach lacks robustness by not taking full advantage of the array of sensors. The beamform- ing algorithm based on steered-response power （SRP） offers strong robust,anti-reverberation, anti-noise, and space filtering capability. It can also enhance Signal Noise Ratio（SNR） , and increase detecting distance. The acoustic localization system relys on a optimization method, such as least-squares（LSQ） algorithm. But LSQ algorithm is easy to be trapped in local opti- mums. Genetie algorithm is a global research method, and can be trapped in global optimums. Gunshot acoustics localization method using SPR and GA algorithm has better precision, proved by field data.