单元机组的小波神经网络广义预测控制  

Wavelet Neural Network Generalized Predictive Control for Generating Unit

作　　者：凌呼君[1] 朱俊峰[1] 李晓明[1] 

机构地区：[1]内蒙古工业大学信息工程学院,内蒙古呼和浩特010080

出　　处：《控制工程》2010年第3期316-319,共4页Control Engineering of China

基　　金：国家自然科学基金资助项目(69964001);内蒙古自然科学基金重点资助项目(200408020802)

摘　　要：针对大型火电机组具有控制对象复杂、非线性、大滞后、模型难以建立等特点,设计协调控制系统的控制策略。小波神经网络具有良好的函数逼近能力和模式分类能力,广义预测控制对比较复杂的工业生产过程呈现良好的控制性能和鲁棒性。通过训练小波神经网对大型火电机组建模得到预测模型,然后利用了模型辨识过程中已获得的数据,计算广义预测控制率,避免了广义预测控制求解丢番图方程带来在线计算量较大的缺陷。仿真结果表明,该方法能保证功率、主蒸汽压力快速平稳地跟踪其设定值,具有较好的鲁棒性,为解决大型单元机组协调控制问题提供了一条有效的途径。A control strategy of the coordinated control system（CCS）is designed,according to the main features of the large thermal-point unit such as control object complex,nonlinear,time-variant,long-time delay and difficulty to establish accurate models,etc.Wavelet neural network（WNN）has a good performance of approximation and classification capacity model,and generalized predictive control has good control performance and robustness in the complex industrial processes.The prediction model is presented to the large thermal-point unit through training WNN.The generalized predictive control rate is calculated by using data in the pattern recognition processes.The defects of a large quantity of calculating generalized predictive control in solving the Diophantine equation.Simulation results show that this way can ensure the power and the main steam pressure in setting value,which provides an effective way to solve the problem of large-scale point unit CCS.

关 键 词：单元机组 小波神经网络 广义预测控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]