Summary: | 碩士 === 南台科技大學 === 電機工程系 === 102 === Comparing with traditional dc motors and induction motors, brushless dc motors
(BLDCMs) provide the merits of no brushes, low consumption, low volume, better
stability and larger torque. As a result, BLDCMs are popular in industrial automation
systems. Generally, the drives with pulse-width modulated inverter are used in the
motor control system. In order to get smooth and stable output torque, the Hall-effect
sensors are employed to sense the rotor position for commutation so that the common
six-step driving algorithm will be applied to the system.
Proportional-Integral-Derivative (PID) control is the widest algorithm used in
control systems. But it is difficult to tune its gains at various conditions. In this thesis,
self-tuning neural network control is considered to assist the PID control for gain tuning
in a short period.
The experimental results are first provided to verify the correctness of self-tuning
neural network plus PID control for 400W-motor control systems in the thesis. Secondly,
an electric vehicle is set up for applications. This electric vehicle includes two sets of
the designed motor systems and CAN BUS transmission that it can move the straight
line and curve paths.
|