Neural-Network-Based Self-Tuning PI Controller forPrecise Motion ...

runmidgeAI and Robotics

Oct 20, 2013 (4 years and 24 days ago)

77 views

控制原理與設計期中報告

指導教授:曾慶耀



號:
10167030



生:楊長諺


Introduction


System Modeling of the PMAC Motor


Neural
-

Network
-

Based Self
-

Tuning PI Control
System for PMAC Motors


Experiments and Discussions


Conclusion


PI control schemes



The artificial neural network technique

A
-
1. Electrical Governing Equation:

A
-
2. Mechanical Governing Equation:

B. Neural
-
Network
-
Based Friction Model

C. Complete Model of the PMAC Motor

D. PMAC Motor PI Control

A. Controller Structure

B. NNPT Training

k
1
=100

k
2
=5

k
3
=100

C. System Integration

D. Computer Simulations

1) Self
-
Tuning PI Control versus Fixed
-
Gain PI Control:

2)

Self
-
Tuning PI Control versus Gain
-
Scheduling PI
Control:





Neural
-
Network
-
Based Self
-
Tuning PI Control:


Neural
-
Network
-
Based Self
-
Tuning PI versus Fixed
-
Gain


In this paper, a new neural
-
network
-
based self
-
tuning
PI controller design method was proposed to increase
the robustness of the conventional fixed
-
gain PI control
scheme.



a well
-
trained neural network supplies the PI controller
with suitable gain according to each feedback operating
condition pair (torque, angular velocity, position error).