基于域变换和灰色预测的光栅信号软细分方法

Soft subdivision method for the grating signal using grey prediction model based on time-space domain transformation

针对在现有的光栅细分方法中,细分精度和细分倍数受光栅输出信号质量制约的问题,提出了一种基于域变换和灰色预测的光栅信号软细分方法。基于时空域变换原理将空间域的信息变换到时间域,将传统的等时间采样转换为等空间采样得到空间序列,然后根据灰色预测理论模型预测代表光栅空间位移信息的时间量,通过模型残差检验和修正算法不断提高预测的准确度,最后以时间脉冲方式输出光栅细分信号。实验研究表明,采用灰度预测模型对光栅信号实现预测的软细分方法,不受信号的正弦性、正交性和等幅性影响,细分误差精度可以达到±1.8″,细分精度优于信号周期的±5%。

In order to improve the subdivision precision and subdivision numbers restrained by the original signal of grating,a soft-subdivision method of grating displacement signal based on domain transformation and grey prediction model is presented. Using the method of time-space transformation,the space domain information is transformed into the time domain information,and those results sampled during equal time intervals are transformed into those sampled during equal spatial intervals. The time quantity representing displacement value of grating sensor is predicted based on grey prediction model. Model residual examination method and correcting algorithm are promoted to improve prediction accuracy. Subdivision pulses of original grating signal are outputted on manner of clock pulses. Experimental results show that the prediction errors are within ± 1. 8″,and the accuracy of subdivision displacement is better than ± 5% of the original grating pitch,which are hardly affected by sinusoidal deviation,bad consistency and orthogonality of grating original signal.