注：北京市自然科学基金资助项目（NO. 4162025）
作者：李东，艾红

单位：北京信息科技大学自动化学院，北京 100192

中图分类号：TP273         

文献标识码：A         

文章编号：1006-883X(2018)12-0025-08

收稿日期：2018-10-19

 

摘要：针对水泥分解炉温度控制难题，利用某水泥厂实际生产的数据在MATLAB系统辨识工具箱中辨识出了分解炉温度和分解炉喂煤量的一阶延时加滞后的系统模型。利用此模型，采用径向基（RBF）神经网络监督控制算法对分解炉温度进行了控制。仿真结果表明，所用算法控制精度高。在此基础上通过改变分解炉温度参考值，模拟不同工况，验证了此控制算法的有效性。与传统PID控制相比，所用算法在分解炉温度参考值发生改变后再次达到稳定的调节时间更短、响应更快。

关键词：水泥分解炉温度；建模；RBF神经网络；监督控制

 

Temperature Monitoring and Control  in Cement Calciners Based on RBF Neural Network

LI Dong, AI Hong

School of Automation, Beijing Information Science and Technology University, Beijing 100192, China

Absrtact: Aiming at the problem of temperature control of cement decomposing furnace, the system model of the first-order delay and lag of the decomposition furnace temperature and the feed quantity of the calciner is identified by using the data from the actual production of a cement plant in the MATLAB System Identification Toolbox. Using this model, the temperature of the calciner is controlled by radial basis function (RBF) neural network monitoring and control algorithm. The simulation results show that the algorithm has high control accuracy. On this basis, the effectiveness of the control algorithm is validated by changing the temperature reference value of the calciner and simulating different working conditions. Compared with the traditional PID control algorithm, the temperature reference value of the calciner is changed to a more stable regulation time and quicker response.

Key words: cement decomposition furnace temperature; modeling; RBF neural network; supervision and control

 

 

备注：2018年 第24卷 第12期