Current position: Home >> Scientific Research >> Paper Publications

一种神经网络分数阶PID控制器的实现

Release Time:2019-03-10  Hits:

Indexed by: Journal Article

Date of Publication: 2014-12-15

Journal: 计算机应用

Included Journals: ISTIC、PKU

Issue: z2

Page Number: 166-168,199

ISSN: 1001-9081

Key Words: 分数阶PID;神经网络;反向传播算法;最优解;自学习

Abstract: 为解决分数阶PID控制器参数难于整定的问题,设计了一种基于神经网络的分数阶PID控制器。通过采用反向传播( BP)神经网络的参数调节策略,可以实现一种五维参数自学习的PID控制器。将分数阶PID控制器数字化,通过BP算法调节神经网络突触权值,经过调整的神经网络输出作为分数阶PID控制器的参数。经过仿真验证,神经网络分数阶PID控制器比传统PID控制器精度提高6倍且控制更加稳定。

Prev One:Influence of tilt angle on eddy current displacement measurement

Next One:A controllable material removal strategy considering force-geometry model in marine propeller five-axis belt grinding