首先，为了去除测量产生的噪声和误差，引入高斯核函数为每个采样点加权；
To filter out the noise and error arising out of various physical measurement processes and limitations of the acquisition technology, a Gaussian weight is assigned to each point acquired.
在此基础上，建立了最小方差损失函数，并结合高斯·牛顿预测误差方法，提出了稳定的，高性能的，在线的复频率直接估计算法。
A cost function is presented, and by applying Gaussian-Newton type recursive prediction error based method, a stable and efficient online frequency estimation algorithm is derived.
应用推荐