上海交通大学学报

上海交通大学学报（自然版） ›› 2012, Vol. 46 ›› Issue (04): 579-583.

• 自动化技术、计算机技术 • 上一篇    下一篇

基于气动人工肌肉的自适应模糊
小脑模型神经网络位置跟踪控制

沈伟1，施光林2   

  1. （1. 上海海洋大学 工程学院，上海 201306； 2. 上海交通大学 机械与动力工程学院，上海 200240）
  • 收稿日期:2012-04-23 出版日期:2012-04-28 发布日期:2012-04-28

Adaptive Fuzzy CMAC Position Tracking Control Based on Pneumatic Artificial Muscle

 SHEN  Wei-1, SHI  Guang-Lin-2   

  1. （1. College of Engineering, Shanghai Ocean University, Shanghai 201306, China；2. School of Mechanical Engineering, Shanghai Jiaotong University, Shanghai 200240, China）
  • Received:2012-04-23 Online:2012-04-28 Published:2012-04-28

摘要： 针对一种气动人工肌肉驱动的弹簧质量位置控制系统，设计了一种自适应模糊小脑模型神经网络(AFCMAC)控制器.离散抗饱和PID (DASPID)并行监督控制设计保证了控制运行初期不会出现较大的跟踪误差和气压波动，使AFCMAC的在线实时学习调整成为可能.在线实时的自适应算法逐步提高了AFCMAC的控制性能，从而最终完全过渡到AFCMAC控制.通过规划AFCMAC的输入空间，保证了AFCMAC对迟滞力和气压波动等不确定因素的感知能力，为实现AFCMAC控制奠定了基础.对DASPID与AFCMAC控制器的位置跟踪控制性能进行了对比实验.结果表明，在非线性系统条件下，AFCMAC较之DASPID有着更好的跟踪控制性能和较低的实现难度.

关键词: 气动人工肌肉, 自适应模糊小脑模型神经网络, 位置跟踪控制

Abstract: A single degree freedom pneumatic artificial muscle spring mass system was built. An adaptive fuzzy CMAC (AFCMAC) was setup to track and control the pneumatic artificial muscle system. With the parallel supervision of discrete anti saturation PID (DASPID) there is few too big tracking error and pressure fluctuation in the beginning of control process, which makes the online real time self adjustment of AFCMAC possible. The online real time self adjustment ability improves the control performance of the adaptive fuzzy CMAC, and finally the spring mass system is entirely controlled by AFCMAC. In the end, the comparative study of control performance between DASPID and CMAC was carried out. The experimental results suggest that in nonlinear condition, AFCMAC could get better tracking performance than DASPID.

Key words: pneumatic artificial muscle, adaptive fuzzy cerebellar model articulation controller(AFCMAC), position tracking control

中图分类号: