基于改进变步长算法的光学电流互感器信号稳定性分析

Signal stability analysis of optical current transformer based on improved variable step size algorithm

为了提高光学电流互感器信号稳定性,构建得到了一种改进后的双SM函数变步长LMS算法,用微软开发的以组件对象模型实现的Active技术建立了具备变步长特征的自适应滤波系统,之后在此系统上比较了不同的变步长算法优劣性。采用SVSLMS算法进行滤波处理得到的输出信号发生了明显失调的情况,形成了很大的稳态误差;处于低信噪比条件下时算法依然存在很高的稳态误差,经过改进处理后算法获得了更快收敛速度并明显降低了稳态误差,从而可以更加精确还原电流信号。相比较其他算法,经过改进后的算法达到更高的信噪比,说明本算法具备更优异的抗噪声干扰性能,使低信噪比条件下的OCT达到更高的稳定性。

In order to improve the stability of optical current transformer signal,got an improved double SM function variable step size LMS algorithm is established,developed by Microsoft to Active technology of component object model established has the characteristics of variable step size adaptive filtering system,on the system after comparing the advantages and disadvantages of different algorithm.SVSLMS algorithm is adopted to filter the output signal,which is obviously maladjusted and forms a large steady-state error.Under the condition of low signal-to-noise ratio,the algorithm still has a high steady-state error.After improved processing,the algorithm obtains a faster convergence speed and significantly reduces the steady-state error,so that the current signal can be more accurately reduced.Compared with other algorithms,the improved algorithm achieves a higher signal-to-noise ratio,which indicates that the algorithm in this paper has better anti-noise performance and makes OCT with low signal-to-noise ratio achieve a higher stability.