摘要
利用BP神经网络对被控对象进行了控制和辨识,提出了一种基于BP神经网络的PID控制器;给出了相应的控制算法;并对典型的参数时变非线性系统的控制进行了仿真研究。仿真结果表明,同传统PID控制器相比,神经网络PID控制器对于模型、环境具有较好的适应能力与较强的鲁棒性,证明了神经网络控制的优越性。
An approach which utilizes the BP neural network for process identification and control is presented. A PID controller based on the proposed neural adaptive control scheme is proposed,and the corresponding algorithm is given. The performance of a typical time-variable nonlinear system is simulated by using the neutral network PID controller. The simulation results show that the neutral net-work controller can improve the robustness of the system and has better adaptability to the model and environment than the classical PID controller.
出处
《合肥工业大学学报（自然科学版）》
CAS
CSCD
北大核心
2006年第11期1375-1379,共5页
Journal of Hefei University of Technology：Natural Science
