The PID controller, which has such a simple structure as to be implemented cheaply, is well-known and found in many practical applications. Also, it has the ability to eliminate steady-state error using integrator and to predict the future behavior of plant using differentiator. PID controller owes these ability to its parameters based on heuristics and some rule of thumbs. But, if the parameters of the PID controller are poorly tuned, its potential cannot be fully utilized. Therefore, in this thesis, an approach to automatically tune and adapt the PID control parameters using the fuzzy logic is presented. The fuzzy logic represents the expert's knowledge to tune and adapt the PID control parameters which are initially tuned by the Ziegler-Nichols method following the relay experiment. The relay experiment makes the plant oscillate and finds the point, the intersection of the transfer function of the plant and the describing function of the relay in the Nyquist plot, then finds the ultimate gain, $k_c$ and the ultimate period, $t_c$. From the $k_c$ and $t_c$, the Ziegler-Nichols method find the parameters of the PID controller, however, the PID controller having parameters determined by the Ziegler-Nichols method may still have large overshoot and long settling time. Then, the proposed method tune well the parameter of PID controller to satisfy the specification given by a user. By applying the proposed method to a simulation and a direct-drive arm, we can find the PID control parameters to satisfy the specification. The proposed method to find the proper parameters of PID controller operates well enough to get the overshoot below 3%.
