基于微分进化的MFFT轧辊偏心诊断技术

摘要/Abstract

摘要：

针对FFT对轧辊偏心信号分析的局限性,提出了基于微分进化的MFFT方法。论述了方法的基本思想、实现过程及其关键步骤。通过仿真实验表明,基于微分进化的MFFT方法可以较好地分析当采样持续时间不是信号周期整数倍时的轧辊偏心信号。以某五机架冷连轧机为研究对象,应用基于微分进化的MFFT方法实现对轧辊偏心的诊断,结果表明:在该五机架连轧机中,第1架轧机的轧辊偏心因素、末架轧机的轧辊偏心因素及热轧前钢坯上的水印因素,是造成末架轧机轧制力波动的主要原因。该诊断方法可以为轧辊偏心的定性分析与在线控制及补偿提供可靠的依据。

关键词:

微分进化, MFFT, 轧辊偏心, 故障诊断

Abstract:

As for the problem of roll eccentricity signals analyzed by FFT,

a MFFT based on the difference evolution was proposed.The basic algorithm,realization process and pertinent programming step were established.The simulation results show that the MFFT method by difference evolution can effectively analyze first-harmonic and harmonic components of eccentricity force signals under conditions that sampling duration

were unequal to the integral multiple of the signal fundamental periodicity.The validation results demonstrate that the eccentricity influence of 1st stand is more serious and that

of 5th mill is second in 5 stand tandem cold rolling mill,further more the effects of skid marks before hot rolling is the primary factor to cyclical force deviations.The prognosis method provides credible the verification on the qualitative analysis and on–line control and compensation of the roll eccentricity.

Key words:

difference evolution, MFFT, roll eccentricity, fault diagnosis

中图分类号:

TP391
TH122

王哲英, 王可, 孙兴伟.

基于微分进化的MFFT轧辊偏心诊断技术

[J]. 中国机械工程, 2010, 21(02): 169-174.

WANG Zhe-Yang, WANG Ge, SUN Xin-Wei-.

Prognosis System for Roll Eccentricity with MFFT Based on the Difference Evolution Algorithm

[J]. China Mechanical Engineering, 2010, 21(02): 169-174.

参考文献

[1]胡广书编著.数字信号处理理论、算法与实现[M],1997:490.
[2]Gareia D F, Lopez J M. A Novel Real- time Fuzzy-- based Diagnostic System of Roll Eccentricity Influence in Finishing Hot Strip Mills[C]//IEEE Industry Applications Society Annual Meeting. New York, 1997:2117-2124.
[3]Huang Min, Fang Xiaoke, Wang Jianhui, et al. Roll Eccentricity Compensation Control for Strip Rolling Mills Based on Wavelet Packet De--noising Theory [C]//Proceedings of the 5th World Congress on Intelligent Control and Automation. Beijing, 2004: 3565-3569.
[4]Wit C D, Praly C L. Adaptive Eccentricity Cornpensation[J]. IEEE Transactions on Control Systems Technology, 2000,8 (5) : 757-766.
[5]Bodson M,Douglas S C. Adaptive Algorithm for the Rejection of Periodic Disturbances with Unknown Frequency[J]. Automatica, 1997, 12 (33): 2213-2221.
[6]Canudas C D,Lischinsky P. Adaptive Friction Corn pensation with Partially Known Dynamic Friction Model[J]. Adapt Control Signal Process, 1997,11 (3) : 65-85.
[7]李育苗.改进的快速富里叶变换及其在轧钢机偏心补偿中的应用[J].自动化学报,1990,16(2):151-155. 0
[8]李育苗.冷轧机中偏心信号补偿的分析与仿真[J].北京科技大学学报,1989,11(3):238-244.
[9]Gu D W, Poon F W. A Robust Fault detection Approach with Application in a Rolling- mill Process[C]//IEEE Transactions on Control Systems Technology. Beijing, 1996 : 408-414.
[10]阳明盛主编.最优化原理、方法及求解软件[M],2006:326.
[11]王哲英 [1],王君 [2],马明旭 [3],王哲锋 [4],刘相华 [2].提高轧辊偏心信号频谱分析精度的方法[J].钢铁,2005,40(8):56-59.
[12]王哲英[1],马明旭[2],朴风贤[3],王哲锋[4].最小二乘法分离频率相近的正弦信号[J].电子与信息学报,2006,28(1):60-65.
[13]Choi S G,Johnson M A,Grimble M J. Polynomial LQG Control of Back- up- roll Eccentricity Gauge Variations in Cold Rolling Mills[J]. Automatica, 1994,30(6) :975-991.
[14]Storn R. On the Usage of Differential Evolution for Function Optimization[C]//Annual Conference of the North Ameriean Fuzzy Information Processing Society-- NAFIPS. New York, 1996 : 519-523.