摘要
为研究基于模式运动的系统动力学描述方法中聚类参数对生产过程调节性能的影响,给出描述系统动态调节性能与产品质量调节性能的指标,分析并建立了聚类参数与系统调节性能间的关系;介绍了基于模式运动的一类复杂生产过程建模方法,并利用LMI方法给出了状态反馈控制器设计方法;提出了基于粒子群优化方法的最大熵聚类算法,定义并提取了系统调节性能指标;利用提出的新的覆盖分类神经网络,建立最大熵聚类方法的参数与调节性能间的映射关系,并分析了分类网络泛化能力;采用实际烧结矿生产数据进行仿真,结果表明所提方法可以分析与建立调节性能与聚类参数间的关系,且可为实际生产中聚类参数的选择提供一定的依据.
In order to study the influence of clustering parameters on the regulation performance of production processes in the pattern-moving-based system dynamics description method,indexes describing dynamic regulation performance and product quality regulation performance are proposed,and the relationship between clustering parameters and regulation performance is analyzed and built.A pattern moving based modeling method of a class of production processes is introduced.A state feedback controller is designed by using the LMI method.A maximum-entropy clustering algorithm based on the particle swarm optimization(PSO)is proposed,and regulation performance of process control is defined and extracted.The relationship between clustering parameters and regulation performance is built by using a proposed classification neural network based on a covering algorithms,and generalization of the network is analyzed.Data of an actual sintering process is used for simulation experiments,and results demonstrate that the proposed method can be used to analyze and build the relationship between clustering parameters and regulation performance,which provides a basis for the selection of clustering parameters in actual production.
作者
徐正光
王目树
郭玲利
XU Zheng-guang;WANG Mu-shu;GUO Ling-li(School of Automation and Electrical Engineering,University of Science and Technology Beijing,Beijing 100083,China)
出处
《控制与决策》
EI
CSCD
北大核心
2020年第5期1025-1038,共14页
Control and Decision
关键词
模式运动
调节性能
神经网络
模式聚类
模式识别
过程控制
pattern moving
regulation performance
neural network
pattern clustering
pattern recognition
process control
