摘要
以加拿大Quanser公司生产的多容水罐为被控对象,研究了基于神经网络模型的非线性预测控制对于该对象的控制效果。采用基于三层静态前向神经网络辨识模型作为这一非线性对象预测模型,利用Matlab/Simulink对双容结构模型进行了仿真研究,证实了神经网络预测控制算法的优点和可行性。
The coupled water tank produced by Quanser company is applied as controlled plant and the neural network nonlinear predictive control is studied.The multilayer feed-forward neural network is adapted as the predictive model,simulation research is done through Matlab / Simulink for the case of double-volume water tank,and the results demonstrated the effectiveness of the proposed scheme.
出处
《工业控制计算机》
2013年第4期70-71,共2页
Industrial Control Computer
关键词
多容水罐
神经网络
非线性预测控制
coupled water tanks
neural network nonlinear predictive control
