论文部分内容阅读
在板材的生产过程中,由于轧辊偏心扰动影响板材厚度的精度,而传统FFT算法在轧辊偏心控制中存在识别误差。针对上述问题,提出了一种基于单子带重构改进算法和比值校正法相结合的偏心信号重构新方法。首先采用单子带重构改进算法对偏心信号进行分频,并且有效的去除噪声,在信噪比低和频率相近时能够准确的提取偏心信号的频率及谐波个数,弥补了比值校正法对噪声敏感的弱点及对各谐波分量间隔大的要求。然后使用比值校正法对各次谐波幅值和相位进一步估计。仿真及实验结果表明,利用结合方法重构得到的轧辊偏心模型可以很大程度减小厚度波动,使轧辊偏心控制精度达到了补偿要求。
In the process of sheet metal production, the accuracy of the sheet thickness is affected by the eccentricity of the roll eccentricity, while the traditional FFT algorithm has the recognition error in the roll eccentricity control. In view of the above problems, a new method of reconstruction of eccentric signals based on the improved single subband reconstruction algorithm and the ratio correction method is proposed. Firstly, the single-band reconstruction algorithm is used to divide the eccentricity signal, and the noise is effectively removed. When the signal-to-noise ratio is low and the frequency is similar, the eccentricity signal frequency and harmonic number can be accurately extracted. Noise-sensitive weaknesses and large harmonic component spacing requirements. Then use the ratio correction method to further estimate the magnitude and phase of each harmonic. The simulation and experimental results show that the roll eccentricity model reconstructed by the combination method can greatly reduce the thickness fluctuation and make the roll eccentricity control accuracy reach the compensation requirement.