一种适用浅海声源的多阵列直接定位方法

针对水下传感器网络系统中的声源定位问题，提出了一种适用于浅海声源的多阵列直接定位方法。该方法根据简正波理论建立阵列接收模型，并根据各阵列接收数据建立联合似然代价函数，最后利用遗传算法进行多维搜索，直接获得声源二维位置估计。仿真结果表明，该方法的声源定位误差小于传统的分步定位方法的定位误差，在较高信噪比时的定位误差稳定在50 m内。该方法通过多阵列联合直接得出声源位置估计，避免了传统分步定位方法中二次估计误差的引入，从而提高了声源定位的精度。

A multi-array based direct positioning algorithm for shallow water sources

Aiming at the problem of multi-source positioning in underwater sensor network system, a multi-array based direct positioning method for shallow sea sources is proposed. In this method, the receiving model of multiple arrays is built based on the normal wave theory, and the joint likelihood cost function is established according to the received data of the multiple arrays. Then, the estimation of two dimensional multi-source positions can directly be obtained by the genetic algorithm. The simulation results show that the source positioning error of this method is smaller than the step by step positioning method, and the positioning error is stable within 50 m at higher signal-to-noise ratio. In this method, the estimation of multi-source positions can be performed by combining multiple arrays, which avoids the second estimation error in the traditional step by step positioning method, thereby to improve the source positioning accuracy.