基于PSO优化神经网络的滚珠丝杠副进给系统自适应滑模控制
首发时间:2018-03-30
摘要:为了提高滚珠丝杠副进给系统的跟踪精度,针对被控对象具有摩擦和未知干扰的特征,采用了基于PSO优化神经网络的自适应滑模控制。因为神经网络的隐含层中心和宽度对网络效果影响不可忽略,所以基于粒子群进行改进,并通过仿真对比,证明改进后收敛速度更快,误差更小。根据滚珠丝杠副的刚体模型,选择未知项利用神经网络进行逼近,并设计具有稳定性的滑模自适应律。实验结果表明:本文设计的控制器跟踪效果较好,相比于PID和优化前方法,可把最大跟踪误差分别从55.05μm和26.53μm缩小至16.85μm。?????
关键词: 机械制造自动化 滚珠丝杠副 神经网络 粒子群算法 自适应滑模控制
For information in English, please click here
Adaptive sliding mode control of ball screw drives based on PSO neural network
Abstract: In order to improve the tracking performance of ball screw drives, because controlled project is affected by the friction and some unknown disturbs, adaptive sliding mode control based on PSO neural network is used. For the non-negligible effect of initial start center and width on neural network, this paper use particle swarm optimization. By contrasting simulation results, the fast rate of convergence and small error are proved. According to the rigid model of ball screw drives, using neutral network to approach the unknown terms. Designing a stable sliding mode controller. The experiment result shows that the tracking performance is good. Compared with PID controller and un-optimized method, the tracking error is respectively reduced from 55.05μm and 26.53μm to 16.85μm. ?????
Keywords: machinery manufacture automation ball screw neutral network particle swarm optimization adaptive sliding mode control ?????
基金:
引用
No.****
动态公开评议
共计0人参与
勘误表
基于PSO优化神经网络的滚珠丝杠副进给系统自适应滑模控制
评论
全部评论0/1000