Summary: | 碩士 === 長庚大學 === 電機工程學系 === 99 === The goal of this thesis is to design a balance controller for a ball and beam system, where an inner-outer loop approach is adopted. In the inner loop, a neural fuzzy controller is considered, whereas a fuzzy control is addressed in the outer loop. In order to improve the control performance, the parameters of the inner loop controller can be on-line tuned by a back propagation algorithm. A CompactRIO based embedded experimental system will be constructed to verify the effectiveness of proposed techniques. In addition to the digital control kernel, the function modules of the experimental setup include the analog input, digital input and DC motor driver. Furthermore, a lookup table is utilized to simplify the computation of a Gaussian function for the purpose of real time implementation. From simulation and experimental results, the proposed control scheme can provide better control responses compared to the traditional fuzzy control.
|