为了提高水下自主机器人(autonomous underwater vehicle,AUV)在海洋执行任务时对深度和航向控制的品质,以及它的抗干扰能力.本文提出了一种新的方法,即通过训练人工神经网络来改变模糊PID(proportional integral derivative)控制的模糊规则和隶属度函数,从而实现更加精确的模糊控制.最后将设计的模糊神经PID控制算法与建立的AUV动力模型相结合.为验证模糊神经PID控制器的有效性,将传统PID、模糊PID控制算法作为对比,同时,人为加入了干扰因素.通过MATLAB/Simulink仿真实验的验证发现,采用模糊神经PID控制器来控制AUV,可以获得更少的反应时间,更好的稳定性,以及更强的抗干扰性,而且控制效果远超其他控制方式.
New AUV Control Research Based on Fuzzy Neural PID
In order to improve the quality of the depth and heading control of the autonomous underwater vehicle(AUV)during its tasks in the sea,as well as its resistance to interference.This paper propose a new approach to more accurate fuzzy control by training artificial neural network pairs to vary the fuzzy rules and affiliation functions of fuzzy PID(proportional integral derivative)control.Finally the designed fuzzy neural PID control algorithm is combined with the established AUV dynamics model.To verify the effectiveness of the fuzzy neural PID controller,conventional PID and fuzzy PID control algorithms were used as comparison algorithms,while interference factors were artificially added.Through the verification of MATLAB/Simulink simulation experiments,the paper found that using a fuzzy neural PID controller to control the AUV,we can obtain less reaction time,better stability,and stronger anti-disturbance,and the control effect far exceeds other control methods.
万方数据
维普
