基于小波神经网络的通用模型控制  

Common Model Control Based on Wavelet Neural Network

作　　者：徐海烽[1] 郭丙君[1] 

机构地区：[1]华东理工大学信息学院,上海200237

出　　处：《化工自动化及仪表》2008年第5期12-15,共4页Control and Instruments in Chemical Industry

摘　　要：针对通用模型控制要求被控对象有显式解的局限性,应用小波神经网络来建立非线性被控对象的逆模型。再结合通用模型控制算法,将非线性过程模型直接嵌入到控制器中,来实现对被控对象的逆控制。其参考轨迹是一条典型的二阶曲线,控制器参数具有明显的物理意义,且易于整定。仿真结果验证了该控制策略的有效性。In order to overcome the weakness that nonlinear controlled objects should have explicit solution in the common model control scheme, used wavelet neural network to set up inverse model of controlled objects. Combined with common model control scheme, nonlinear process model could embed directly into the controller to realize inverse control. The reference trajectory is a classic second order curve. The controller parameters have clear physical mean- ing and are easy to tune. The simulation results show the effectiveness of the proposed control scheme.

关 键 词：通用模型控制 小波神经网络 逆控制 非线性 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]