基于粒子群理论的板形模糊模式识别方法  

Fuzzy Pattern Recognition Method of Flatness Based on Particle Swarm Theory

作　　者：刘建[1] 王益群[1] 孙福[1] 宁淑荣[1] 

机构地区：[1]燕山大学机械工程学院,秦皇岛066004

出　　处：《机械工程学报》2008年第1期173-178,共6页Journal of Mechanical Engineering

基　　金：国家自然科学基金(60474044;50675189);河北省自然科学基金(E2004000221)资助项目。

摘　　要：带钢板形的模式识别是板形闭环控制的关键环节,板形模式的识别结果直接影响着板形控制精度。随着板形控制手段的不断更新,对板形模式识别方法提出更高的要求。为克服传统板形模式识别方法抗干扰能力差、逼近阶难以确定的缺点,依据模糊分类原理,运用欧式距离的择近原则对板形模式进行分类,完成板形信号的模式识别。在此基础上,为进一步提高识别精度,将20世纪90年代发展起来的具有全局优化能力的粒子群理论应用于板形模式识别,对模式识别的结果进行优化,并将其与单纯形法优化结果进行对比。试验结果证明了粒子群优化算法的有效性,该算法能够提高识别精度,使优化后的结果能更精确地控制板形调控机构,以适应高精度板形控制要求。Pattern recognition of strip flatness is the key to flatness close-loop control system. The result of pattern recognition of strip flatness has the direct effects on strip flatness control precision. With the development of strip flatness control methods, higher demands are presented for pattern recognition of strip flatness. To overcome the disadvantage of poor anti-interference ability and uncertain approaching ranks in traditional flatness pattern recognition, the fuzzy pattern recognition method of flatness is given according to the fuzzy classification theory, and the flatness pattern is classified by selecting nearness principle of Euclidean distance. On the basis of fuzzy pattern recognition, particle swarm theory, developed since 1990s, with global optimization capability is applied to optimize the result of pattern recognition. Compared with the simplex optimization method, validity of particle swarm theory applied in flatness pattern recognition is testified. The result after optimization can accurately control the flatness adjusting sets to meet the need of high precision flatness control.

关 键 词：板形 模糊模式识别 粒子群 欧式距离 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]