基于混沌优化的非线性预测控制器

针对非线性系统的控制问题,本文将神经网络辨识、混沌优化和预测控制思想有机结合,提出了一种新型非线性预测控制器.该控制器以神经网络作为预测模型,混沌优化算法作为滚动优化策略,避免了非线性预测控制中复杂的梯度计算和矩阵求逆问题.另外在训练神经网络过程中,采用了带混沌机制的自适应学习率的BP算法,以提高神经网络的收敛能力和收敛速度.仿真研究说明了该非线性预测控制器的有效性及实时性.

A nonlinear predictive controller based on chaos optimization

A new nonlinear predictive controller is proposed which combines neural network identification,chaos opti- mization algorithm(COA)and the concept of predictive control.The controller utilizes neural network as predictive model and COA as online optimization.It can avoid calculating the complicated gradient and the inverse matrix in the nonlinear predictive control.For training the neural network,moreover,chaotic mechanism and adaptive learning rate are adopted into the normal backpropagation(BP)algorithm to improve the network convergence.The simulation studies show the effective performance of the proposed controller.