MVDR自适应波束形成的一种智能算法

最小方差无失真响应（MVDR）自适应波束形成方法在声纳阵列信号处理中得到了广泛的应用。当存在强干扰时,传统的MVDR算法的稳定性较差,在有限次快拍数条件下,会带来一定的信噪比损失。将粒子群优化算法（PSO）引入MVDR的求解过程,提出了一种MVDR的数值实现方法。该方法在保持信号能量不变、干扰和噪声能量最小的约束条件下,利用粒子群优化算法通过数值搜索的方法获得MVDR的最优权向量。在小快拍数条件下,较一般的MVDR具有更好的抑制干扰能力。计算机仿真结果验证了该方法的有效性。

An intelligent algorithm of MVDR adaptive beamforming

MVDR（Minimum Variance Distortionless Response） adaptive beamforming method is widely used in sonar array signal processing.The conventional method has poor performance and low SNR gain in the condition of limited snapshots or strong interference signal existing.A numerical realization method of MVDR is proposed in this paper.The proposed algorithm is realized via numerical search by particle swarm optimization method which maintain the signal and minimize the power of noise and interference in the constrain condition.Compared to conventional MVDR beamforming algorithm,the proposed algorithm has better performance in suppressing the interference and noise when the snapshots are limited,which is confirmed by computer simulations.