论文部分内容阅读
锅炉主蒸汽温对象的复杂非线性动态特性使得常规设计的固定参数比例-积分-微分(PID)串级控制很难适应工况变化和保证各种环境下的控制品质。为此,采用基于SOM的RBF-ARX模型框架,通过运行数据离线辨识的方法获取主汽温对象全局动态模型参数,在该模型基础上设计具有滚动优化和反馈校正功能的广义预测控制(GPC)器。220t/h高压煤粉锅炉模型上的仿真结果和130t/h煤粉炉上的工程应用均显示该方法得到的全局动态模型具有较高的预测精度,设计的GPC控制器具有较强的工况变化适应能力,能够适应多煤种变化和负荷变化。
The complex non-linear dynamic characteristics of the main steam temperature of the boiler make it difficult for the conventional design of proportional-integral-derivative (PID) cascade control to adapt to changes in conditions and to ensure the quality of control in various environments. Therefore, based on SOM-based RBF-ARX model framework, global dynamic model parameters of main steam temperature object are acquired by offline identification of running data, and generalized predictive control (GPC) with rolling optimization and feedback correction is designed based on this model. Device. The simulation results on the 220t / h high pressure pulverized coal boiler and the engineering application on the 130t / h pulverized coal fired boiler show that the global dynamic model obtained by this method has high prediction accuracy. The designed GPC controller has strong working conditions Adaptability to change, to adapt to changes in coal and load changes.