摘要
针对线性PID控制器系数难以整定的问题,构造了一种用神经网络实现的非线性PID控制器．多个具有相同结构的非线性PID控制器并联,对多变量系统实现解耦控制器．结合预测控制的思想,提出两种控制方案．第一种是在递归多步预测的基础上,在广义最小方差目标函数下实现控制,第二种利用多步预测目标函数在线修正解耦控制器的权值．仿真实验表明这两种方法的有效性．
A nonlinear PID controller is proposed based on recurrent neural network, due to the difficulty of tuning the parameters of conventional PID controller. In the control process of nonlinear multivariable system, a decoupling controller is constructed, which takes advantage of multi-nonlinear PID controllers in parallel. Under the idea of predictive control, two multivariable predictive control strategies are established. One is that the general minimum variance control function is used based on recursive multi-step predictive method. The other is that the multi-step predictive cost energy is adopted to train the weights of the decoupling controller, Simulation studies have shown their efficiency.
出处
《南开大学学报（自然科学版）》
CAS
CSCD
北大核心
2006年第1期49-53,61,共6页
Acta Scientiarum Naturalium Universitatis Nankaiensis
基金
国家自然科学基金(60374037)
国家科技攻关计划(2004BA204B08-02)
南开大学科技创新基金
关键词
预测控制
解耦控制
递归神经网络
非线性PID控制
predictive control
decoupling control
recurrent neural networks
nonlinear PID control
