论文部分内容阅读
在水声信号盲分离技术仿真研究中 ,需要评价算法的收敛速度 ,稳态误差等。文中分析了水声基阵接收信号的实数延时模型和复数混合模型 ,提出了水声信号盲分离算法性能评价准则。该准则分为基于混合基阵和信号估计两种。在三 /四元均匀线列阵上 ,盲分离实测非高斯舰船辐射噪声 ,分别用 CRLS、CACY、EASI、CMA、MKMA算法验证了提出的算法评价准则 ,同时 ,还给出了这些算法分离水声信号的性能曲线。
In the simulation of underwater acoustic signal blind separation technology, we need to evaluate the convergence speed, steady-state error and so on. In this paper, the real delay model and complex mixing model of underwater acoustic array receiving signals are analyzed, and the performance evaluation criteria of underwater acoustic signal blind separation algorithm is proposed. The guidelines are divided into two categories based on mixed-matrix and signal estimates. In the three-and four-element uniform linear array, the radiation noise of non-Gaussian warship was measured by blind separation. The CRLS, CACY, EASI, CMA and MKMA algorithms were respectively used to verify the proposed algorithm evaluation criteria. At the same time, Acoustic signal performance curve.