Power System Stabilizer Design Using Adaptive FOPID Controller based on Self-Learning Wavelet Neural Networks

Several methods were proposed for the design of power system stabilizer, PSS, based on PI, PID, and FOPID controllers. In these controllers, the degree of freedom increases from two to three and five, respectively. Although increasing the degree of freedom can enhance the convergence rate and the robustness of the controller, it does come with more challenges when it comes to tuning the control parameters. For instance, it is no longer possible to adjust FOPID parameters using trial and error. One of the conventional methods is to use optimization algorithms, but it should be noted that the power system is highly non-linear. This research aimed to propose an algorithm to design the PSS controller based on FOPID, in which the controller coefficients were adjusted based on the system conditions. For this purpose, the controller coefficients were defined based on the gradient of the power system, so that the coefficients were adjusted at any moment by the adaptive-indirect gradient method in such a way that the cost function of the controller was minimized, and as a result, the rate of oscillation damping increased. In the proposed algorithm, an identifier based on self-tuning wavelet neural network with online learning was used to estimate the gradient of the power system. Finally, the proposed adaptive controller was designed for a two-zone, two-machine power system including FACTs devices, SSSC-type, and its performance was evaluated in comparison with other methods. The results confirm the effectiveness of the proposed method.