控制器参数仅在与采样周期成倍数关系的时刻改变，避免了控制器频繁而剧烈的调整。
The controller parameters are updated only at sampling instant that is equal to multiple cycle, so continous and rapid controller adjustment is avoided.
采用正切细分法，对不足一个信号周期的条纹位移，实现了正确的辨向和高倍数细分。
Combining with tangent subdividing method, this method can correct the counting error, recognize the direction and subdivide the fraction of fringes displacement.
由于进动周期的上下限之比通常大于2，因此该方法的计算量比自相关法增大的倍数小于1.4；同时它对先验信息的要求非常宽松。
Usually the ratio of the upper bound to the lower bound is larger than 2, resulting in the computational burden of the proposed method less than 1.4 times of that of the autocorrelation method.
应用推荐