【摘要】 针对目前大多数独立分量分析(independent component analysis,ICA)算法无噪或者弱噪声假设的局限性,提出一种适用于低信噪比情况的有噪独立分量分析算法。该算法以分离信号的负熵为目标函数,采用高斯分布密度模型作为非线性函数来估计负熵,并建立了模型参数的确定准则,能够较好地抑制低信噪比下噪声的影响,最后采用人工蜂群算法对混合矩阵进行全局寻优。仿真结果表明,与其他算法相比,提出的算法可以更为精确地估计混合矩阵,能够较好地解决低信噪比下的有噪ICA问题。更多还原

【Abstract】 Aiming at the limitation that most ICA algorithms are based on the noise-free or weak noise model, a noisy independent component analysis algorithm which is suitable for low signal-noise ratio is proposed. The negentropy of the separated signals is utilized as the objective function. The Gaussian-distribution density model is selected as the nonlinear function to estimate the negentropy. Meanwhile, a confirm rule of the model parameters is established, which can suppress the noise effects in low SNR. Finally, the mixing matrix is optimized by the artificial bee colony algorithm. Simulation results illustrate that compared with other algorithms, the proposed method can estimate the mixing matrix parameters more accurately and solve the noisy ICA problems better.更多还原