焦炉冷鼓系统自适应PID控制器设计  

Design of Adaptive PID Controller for Coke Oven Cooling Blower Drum System

作　　者：关慧敏[1] 张世峰[1] 董鑫[1] 许四长 GUAN Huimin;ZHANG Shifeng;DONG Xin;XU Sichang(School of Electrical and Information Engineering, Anhui University of Technology, Ma'anshan 243032, China)

机构地区：[1]安徽工业大学电气与信息工程学院,安徽马鞍山243032

出　　处：《安徽工业大学学报（自然科学版）》2016年第2期142-147,共6页Journal of Anhui University of Technology（Natural Science）

基　　金：安徽省教育厅自然科学研究项目(KJ2008B104)

摘　　要：焦炉冷鼓系统的稳定状态直接影响炼焦生产的工艺指标。根据现场不同工况(正常工况、检修保温工况、非正常工况),采用最近邻聚类学习算法训练的RBF网络辨识,建立焦炉冷鼓系统及其控制系统的仿真模型,辨识出Jacobian信息并用于BP神经网络整定PID参数,实现适应不同工况的自整定PID控制。仿真结果表明,建立的模型及控制系统能将冷鼓系统的初冷器前吸力快速、有效地稳定在一定范围内,控制精度高、稳定性好,可保证焦炉冷鼓系统在不同工况下稳定运行,其自适应能力对稳定生产工艺指标具有一定的有效性。Steady state of coke oven cooling blower system directly affects process parameters of cokingproduction. According to three different conditions (normal operating conditions, maintenance insulation conditions,abnormal conditions) on site, the nearest neighbor clustering algorithm was employed to train RBF neuralnetwork identification, and the simulation models of coke oven cooling blower system and its control system wereestablished, the identified Jacobian information was used for BP neural network to tune PID parameters, andself- tuning PID control under different conditions was implemented. Simulation results show that with theestablished model and its control system the primary cooler anterior suction of cooling blower system can becontrolled within a certain range quickly and effectively, with high control accuracy and good stability, which canensure the stable operation of different condition, and its adaptability has a certain validity in stabilizing theproduction process parameters.

关 键 词：焦炉冷鼓系统 最近邻聚类学习算法 RBF网络 BP神经网络 自适应PID 

分 类 号：TP229[自动化与计算机技术—检测技术与自动化装置]