估计声信号二维参数的免疫量子克隆算法

Immune Quantum-Clonal Algorithm for Estimating Two-Dimensional Parameters of Sound Signals

信号相位匹配(SPM)原理同时估计信号方位和频率两参数时,因计算量大,需设置固定的搜索步长和频带范围。针对该问题,提出了一种利用免疫量子克隆算法(IQCA)对声源波达方向和频率二维参数进行快速联合估计的高分辨方法。将SPM原理的奇异值分解(SVDSPM)判别准则作为具有双自变量(方位和频率)的适应度函数,并通过IQCA对该适应度函数进行非线性全局寻优,不仅实现了SPM的二维联合估计方法,而且加速了搜索最优参数的收敛过程。IQCA算法对SVDSPM的估计不仅提高了算法实时性,且与标准遗传算法(SGA)、免疫遗传算法(IGA)和量子进化算法(QEA)相比,大大提高了估计精度和稳定性。同时,双源估计的结果也表明本文算法在低信噪比下的性能优于MUSIC算法。

The signal phase matching （SPM） principle is difficult to give out the fixed steplength and the frequency band when it searches the direction-of-arrival （DOA） and the frequency of signals simultaneously. Aimed at the problem, a rapid method for joint estimation of two-dimensional parameters with high resolution is proposed by using the immune quantum- clonal algorithm （IQCA）. The novel method constructs the fitness function from the singular value decomposition of the SPM principle （SVDSPM） and takes the DOA and the frequency as two independent variables. The IQCA optimizes the fitness function nonlinearly and globally, realizes the method of the two-dimensional joint estimation based on the SPM, and accelerates the searching convergence of the optimal parameters. Compared with the simple genetic algo- rithm （SGA）, the immune genetic algorithm （IGA） and the quantum evolution algorithm （QEA）, the proposed algorithm improves the accuracy and the＇stability of the parameter esti- mation. Furthermore, the estimation results of two sources show that the performance of the new algorithm is better than that of MUSIC algorithm at the low SNR.