论文部分内容阅读
伯格反褶积又称最大熵反褶积,它是一种高分辨率反褶积方法。其特点是假设时窗外的数据具有最大的随机性。它与预测反褶积方法的主要区别在于不是根据输入信号的自相关函数求出预测误差算子的全部项,而是通过第m阶预测误差求出第m+1阶预测误差。当预测步长为1时,预测反褶积的输出相当于最小平方脉冲反褶积的输出。
Burg Deconvolution, also known as maximum entropy deconvolution, is a high-resolution deconvolution method. It is characterized by the assumption that the data outside the time window has the greatest randomness. The main difference between this method and the predictive deconvolution method is that instead of calculating all the terms of the prediction error operator based on the autocorrelation function of the input signal, the (m + 1) th prediction error is obtained by using the mth prediction error. When the prediction step is 1, the output of the predicted deconvolution corresponds to the output of the least square pulse deconvolution.