Voice coil motor (VCM) is widely known as its topquality
of free friction, low noise, fast transient response and well
repeatability. Yet the dynamic characteristic of a VCM is
nonlinear and time-varying, thus the model-based conventional
controller is difficult to achieve high-precision control
performance for a VCM. To attack this problem, a selforganizing
fuzzy sliding-mode control (SFSC) system is proposed
in this paper. All of the fuzzy rules are online grown and pruned
by the structure learning phase and the parameter learning
phase is designed to tune the controller parameter in the
gradient-descent-learning algorithm. From the experiment
results, it shows that the proposed SFSC system can successfully
control a VCM with favorable control response with enhanced
disturbance rejection performance.
關聯:
12th IEEE International Conference on Networking, Sensing and Control, pp.287-292