计算机工程与应用 ›› 2022, Vol. 58 ›› Issue (15): 317-323.DOI: 10.3778/j.issn.1002-8331.2201-0413

• 工程与应用 • 上一篇    下一篇

基于神经网络和模糊补偿的水下机械臂控制

高阳,张晓晖,高玉儿,尚婷,杨启航   

  1. 西安理工大学 自动化与信息工程学院,西安 710048
  • 出版日期:2022-08-01 发布日期:2022-08-01

Control of Underwater Manipulator Based on Neural Network and Fuzzy Compensation

GAO Yang, ZHANG Xiaohui, GAO Yu'er, SHANG Ting, YANG Qihang   

  1. School of Automation and Information Engineering, Xi'an University of Technology, Xi'an 710048, China
  • Online:2022-08-01 Published:2022-08-01

摘要: 针对水下机械臂动力学模型建模复杂且滑模控制的抖振问题,利用Lagrange法和Morison方程精准建立二连杆串联水下机械臂的动力学模型,对模型中参数的不确定项使用4个RBF神经网络分别进行逼近,并且对摩擦项使用模糊控制进行补偿的方法,精准迅速地实现了对水下机械臂控制系统跟踪控制。通过进行仿真分析,基于神经网络和模糊补偿控制的方法与滑模控制、整体RBF神经网络控制和分块RBF神经网络控制相比,控制系统的平均误差分别降低了85.5%、71.8%、93.1%。结果表明,此方法有效降低了控制系统的跟踪误差,并同时提高了稳态性和抗干扰性。

关键词: 水下机械臂, RBF神经网络, 模糊补偿, 动力学模型

Abstract: Aiming at the complex modeling of underwater manipulator dynamic model and the chattering problem of sliding mode control, the dynamic model of two-link series underwater manipulator is accurately established by using Lagrange method and Morison equation. The uncertainties of parameters in the model are respectively approximated by four RBF neural networks, and the friction term is compensated by fuzzy control, which accurately and quickly realizes the tracking control of the underwater manipulator control system. Through simulation analysis, the average error of the method based on neural network and fuzzy compensation control is reduced by 85.5%, 71.8% and 93.1% respectively compared with sliding mode control, overall RBF neural network control and block RBF neural network control. The results show that this method effectively reduces the tracking error of the control system, and improves the stability and anti-interference at the same time.

Key words: underwater manipulator, RBF neural network, fuzzy compensation, dynamics model