摘要
在分析系统轮廓误差的基础上,提出了基于模糊神经网络的轮廓误差补偿方法,并说明其补偿器的原理、算法及实现。该法在不改变系统各单轴位置环的前提下,根据系统的轮廓误差,通过模糊神经网络的自学习能力动态向各轴提供误差补偿信息,进而提高系统的轮廓精度,同时也解决了各轴之间增益不匹配、动态不匹配和各轴不可预见性问题。最后,在MATLAB6.1环境下对该系统进行仿真,仿真结果表明其可行性和有效性。
This paper analyses the system contour error and proposes a new contour error compensation method based on fuzzy neural network. It also explains principle, algorithm and realization of this compensator. Through the adaptive ability of fuzzy neural network, this method can dynamically provide each axis with additional compensating information according to system contour error, and then the system contour precision is improved without changing original position controller of each axis. At the same time, gain unmatchable & dynamic characteristic unmatchable & each axis unforeseeable problems are solved. Finally, this system is simulated in MATLAB6.1 and the results indicate that the system is feasible and valid.
出处
《系统仿真学报》
CAS
CSCD
2003年第12期1733-1736,共4页
Journal of System Simulation
关键词
轮廓误差
模糊神经网络
动态分配
补偿器
contour error
fuzzy neural network
dynamic distribution
compensator