| Summary: | 碩士 === 國立臺灣科技大學 === 機械工程系 === 99 === Linear Piezoelectric Ceramic Motors (LPCM) have been recognized as one kind of useful actuators in recent mechatronics and automation industries. Due to their dominant nonlinear effects such as friction and hysteresis, controller design for LPCM has become a challenging task for researchers. The current literature related to motion control of LPCM shows that the designers mostly suffer from either complicated friction modeling for compensation or numerous parameters tuning by using intelligent control algorithms. Following the idea of “add-ons” controller design, this thesis proposes a novel hybrid control structure, comprising of PID control, repetitive control, and neural network adaptive control, to achieve high performance and easy maintainability for motion control of LPCM. The experimental results on tracking motion control of a LPCM demonstrate the effectiveness of the proposed method. Moreover, the study also adds a LuGre friction model based feedforward compensator to the hybrid controller for comparison purposes and further performance improvement. Biaxial motion control experiments are finally provided to justify the applicability of this method.
|
|---|
