摘要
针对潜艇定深运动过程中存在非线性、时变参数、复杂干扰的特点,提出了一种基于神经网络的自适应模糊控制器,并采用学习速率自调整的EBP算法对模糊控制器进行了在线调整.仿真结果表明,该控制器能辨别出潜艇的平衡舵角,与常规的PID控制相比,具有抗干扰能力强、响应速度快、精度高等优点.
During the course of the movement of the submarine in the fixed depth, there are some characteristics of nonlinearity, time-varying parameters and influence of complex disturbance. A kind of fuzzy adaptive controller based on neural networks is proposed. This fuzzy controller is adjusted on-line by EBP algorithm whose learning algorithm is adjusted by itself. The simulation results show that the controller can recognize the equilibrium rudder angle of the submarine and has the advantages of strong ability of anti-interference, quick response speed and high accuracy, while compared with the PID controller.
出处
《海军工程大学学报》
CAS
2004年第2期83-88,共6页
Journal of Naval University of Engineering
