一种变速率非线性盲源分离算法

A Variable Rate Nonlinear Blind Source Separation

基于核函数的非线性盲源分离算法在盲信号处理中有着广泛的应用,但传统的非线性盲源分离算法的学习速率是固定的。当系统的噪声和迭代误差较小而步长相对较小,则算法达到收敛的效率不高;当系统的噪声和迭代误差较大时,如果迭代步长过大则将会影响盲源分离的精度。针对这一不足,提出一种基于核函数的变速率非线性盲源分离算法,算法根据信噪比和迭代误差来调节学习速率,将该算法应用于齿轮箱故障诊断中。仿真和实验结果表明,与固定速率的非线性盲源分离算法相比,该算法具有更好的消噪和信号特征提取能力。

The nonlinear blind source separation algorithm based on kernel function has been widely applied in blind signal processing.However the learning rate of traditional kernel function method is fixed.When the system noise level and iterative error are large,it will take a long time for iterative parameters to get convergence.So it will affect the efficiency of blind source separation.In this paper,a variable rate nonlinear blind source separation algorithm based on kernel function was proposed.The learning rate of the algorithm was adjusted according to SNR and iterative error.The algorithm was applied in gear faults diagnosis.Both of the results of simulation and experiment indicate that the present algorithm has better ability of noise reduction and feature extraction than fixed rate nonlinear blind source separation.