China Mechanical Engineering ›› 2013, Vol. 24 ›› Issue (08): 1023-1028.

Previous Articles     Next Articles

Inverse-model Control for Free-floating Space Flexible Robotic Manipulators Based on Neural Network

Hu Xiaoping;Zhang Wenhui;Ji Xiaoming   

  1. Lishui University,Lishui,Zhejiang,323000
  • Online:2013-04-25 Published:2013-05-08
  • Supported by:
     
    Zhejiang Provincial Natural Science Foundation of China(No. LZ12F02001)

漂浮基空间柔性机械臂基于神经网络的逆模控制

胡小平;张文辉;季晓明   

  1. 丽水学院,丽水,323000
  • 基金资助:
    浙江省自然科学基金资助重点项目(LZ12F02001);丽水市科技局公益应用项目(2012JYZB30);丽水学院校级科研项目(KY201111) 
    Zhejiang Provincial Natural Science Foundation of China(No. LZ12F02001)

Abstract:

Trajectory tracking problems of free-floating space flexible manipulators were studied.A neural network inverse-model control algorithms based on improved Kalman filtering algorithm was proposed herein.A
nonlinear dynamics model was established,and then its control law based on the augmented variable input method was designed;feedforward neural network with good approximation ability was used to compensate adaptivety the unknown nonlinear inverse model. Kalman filtering algorithm was designed to ensure network weights online real-time adjustment,system error function was provided by PID controller.Simulation results show that the proposed control scheme is effective and has high value of engineering applications.

Key words: neural network, Kalman filtering algorithm, flexible manipulator, inverse-model control, trajectory tracking

摘要:

针对传统控制方法对强耦合柔性空间机械臂难以有效控制的问题,提出基于神经网络的逆模控制策略。建立了非线性空间柔性机器人的动力学模型,根据增广变量输入法推得其控制律;利用具有良好逼近能力的前馈神经网络来自适应补偿柔性臂的未知非线性逆模型;采用Kalman滤波算法来保证网络权值在线实时调整(系统的误差代价函数由PID控制器提供)。仿真证明了所提出的控制方案的有效性,具有较高工程应用价值。

关键词: 神经网络, Kalman滤波算法, 柔性机械臂, 逆模控制, 轨迹跟踪

CLC Number: