In this paper, an optimal parameter set tuning method for proportional-integral-derivative (PID) controller and fractional order PID controller is proposed using teaching-learning based optimization (TLBO) algorithm. During the optimization of the PID and FOPID controller parameters, an objective function consisting of overshoot, rise time, settling time and steady state error is formulated to achieve a satisfactory trade-off between the dynamic response characteristics. TLBO algorithm is used as the optimizer to find the best parameters of the proposed controllers. The designed PID and FOPID controllers are applied to an automatic voltage regulator (AVR) system. The performances of the proposed controllers are confirmed by comparing their results with those obtained from different optimized PID and FOPID controllers previously published in the literature for the same AVR system. The numerical simulations and comparisons show that the proposed controllers provide the better dynamic response characteristics as well as more robust to model uncertainties than the other different optimized controllers. The results obtained with the proposed controllers show a better trade-off between the set point tracking performance, robustness and stability.

Issue of balance in robotics is best represented by the balancing act of the inverted pendulums. With their unstable and nonlinear behaviour, Inverted Pendulum systems are quite popular systems in which various control design methods are applied and performance comparisons are carried out. In this study, in order to control the pendulum angle and car position of the inverted pendulum system, a fractional order PID controller is designed. Gains of the designed controller are optimized by Particle Swarm Optimization method. On the other hand, for the aim of comparison of performance, conventional PID controllers are used for controllers. The optimum values of gains for PID controllers are also found by same optimization algorithm. The simulation results of both controllers are compared for the inverted pendulum.