Optimization Study and Applications of Artificial Intelligence Algorithms in the Mechatronic Systems

• Main Authors: Chih-Cheng Kao, 高志成
• Other Authors: Rong-Fong Fung
• Format: Others
• Language: en_US
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/38189654670805868298

博士 === 國立高雄第一科技大學 === 工程科技研究所 === 96 === Artificial intelligence is once of the most popular optimization algorithms. This dissertation focuses on developing a hybrid intelligent controller, based on optimization and optimal control theory. Several solution searching strategies and mechanisms are proposed to improve solution searching capability and efficiency of population-based optimizers for dealing with different types of optimization problems. To verify the solution searching ability in the solution space, the proposed algorithm will be applied to mechanism systems. Some hybrid intelligent controllers are designed for three mechanism systems. Firstly, by using the particle swarm optimization (PSO) algorithm, a novel design method is proposed for the self-tuning proportional-integral-derivative (PID) control in a slider-crank mechanism system. The proposed approach has superior features, including: easy implementation; stable convergence characteristics; and good computational efficiency. Fast tuning of optimum PID controller parameters yields high-quality solutions. Secondly, we propose an effective method for singularity control of a fully parallel robot manipulator using the PSO and grey prediction (GP). By employing to employ the PSO and GP methods to efficiently search the optimal damping values on-line, the motion of the manipulator around the singular point can be controlled effectively, and highly improve the accuracy of system responses. Thirdly, we proposed to utilize Bouc-Wen hysteresis model, to identify a piezoelectric actuator (PA), which drives a Scott-Russell (SR) magnification mechanism. In system identification, we adopt the PSO to find the parameters of the SR mechanism and the PA. All the proposed control schemes are applied to these mechanisms by simulation and experiment to demonstrate the effectiveness and advantages.