基于改进型Elman网络的锅炉预测控制算法研究  

Research on predictive control algorithm based on improved Elman network in boiler

作　　者：费红举[1] 张广明[1] 邱春玲[1] 

机构地区：[1]南京工业大学自动化学院,江苏南京210009

出　　处：《计算机工程与设计》2008年第15期3987-3988,4063,共3页Computer Engineering and Design

摘　　要：针对锅炉燃烧系统非线性、多变量、大延时、强耦合等特点,将预测控制和改进型Elman神经网络相结合,提出了一种基于改进型Elman网络的预测控制算法。该算法可以克服传统PID控制方法对多变量对象控制效果差、抗干扰能力弱等缺点。选取链条炉燃烧系统作为研究对象,仿真结果表明:该算法具良好的控制效果和较强的抗干扰能力。In accordance with the characteristics such as nonlinearity, multivariable, large time delay and coupling of POller burning system, predictive control and improved Elman network are combined, a predictive control algorithm based on improved Elman network is proposed. It can overcome the disadvantages of bad control effect in multivariable object and low anji-jamming ability of traditional control method of PID. The burning system of chain boiler is researched. It can be included from the simulation that the algorithm has good controlling result and strong anti-jamming performance.

关 键 词：锅炉 燃烧系统 预测控制 ELMAN网络 算法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]