焦炉加热燃烧火道温度优化控制仿真研究

摘要

The quality of flue temperature controlaffects the coke quality and the energy consumption directly.Aiming at the problems of high energy consumption of the coking process,low flue temperature control accuracy,bad optimization effect and unreasonable utilization of large amount of production data,the article combines the characteristics of coke oven heating processes such as large lag and strong coupling,and presents an optimal control method of coke oven flue temperature based on Heuristic Dynamic Programming (HDP) for the first time.Using the method of data-driven combined with nonlinear modeling,we adopted the modeling method based on data,established the coke oven flue temperature system model network,critic network and executive network by BP neural network.Simulation was carried out based on MATLAB.Thesimulation results show that the HDPhas good optimal controleffectof flue temperature,and manifold pressure,ammonia pressure,flow coke oven gas,coke oven flue gas mains pressure and temperature can be stabilized in a short time.%焦炉火道温度控制的好与坏直接影响炼焦质量和能耗,针对炼焦过程能耗高,火道温度控制精度不高,优化效果差等问题,以及生产过程中大量有用数据未被合理利用的情况,结合其加热过程强耦合、大滞后等特点,首次提出了基于启发式动态规划(HDP)算法的焦炉火道温度的优化控制方法,采用数据驱动结合非线性建模的方法,使用BP神经网络分别建立了焦炉火道温度系统模型网络,评价网络和执行网络,并在MATLAB环境下进行仿真.仿真结果表明,HDP控制器能够实现对焦炉火道温度的优化控制,且集气管压力、氨水压力、焦炉煤气流量、焦炉煤气主管压力和烟道温度都能在短时间内稳定.提高了火道温度的稳定性,减少了能耗,能更好地指导生产.