基于神经网络模型的最小预测误差非线性自适应控制器
Minimum Prediction Error Adaptive Controller Based on Neural Network Model
-
摘要: 主要研究基于神经网络模型的最小预测误差非线性自适应控制算法.利用神经网 络激励函数的分段局部线性近似,将基于神经网络的非线性系统一步前向预测控制转化为一 系列局部的线性预测控制问题.利用线性系统参数估计方法获得神经网络预测模型的参数估 计.在此基础上利用并联线性系统的预测控制方法设计全局收敛的非线性系统预测控制器.Abstract: The paper mainly investigates the minimum prediction error adaptive controller based on the neural network model. With the help of local linearization of activation function, complicated nonlinear adaptive controller is transformed into an adaptive controller of a series of linear systems in parallel form. Hence, the parameter estimation and control law are straightforwardly implemented with the linear prediction control algorithm.
计量
- 文章访问数: 2287
- HTML全文浏览量: 136
- PDF下载量: 896
- 被引次数: 0