Rank Exponent Method Based Optimal Control of AGC for Two-Area Interconnected Power Systems
Mamta,V. P. Singh,A. V. Waghmare,V. P. Meena,Francesco Benedetto,T. Varshney,Akanksha V. Waghmare,Veerpratap P. Meena,Tarun Varshney
DOI: https://doi.org/10.1109/access.2024.3373043
IF: 3.9
2024-03-12
IEEE Access
Abstract:Automatic generation control (AGC) is employed in power systems to maintain balance between generation and load by adjusting output of generators in real time. Controller continuously monitors system frequency and tie-line power flow by responding to fluctuations in electricity demand and supply and optimizes generator dispatch, reduces power imbalances, and enhances grid stability. This work proposes and solves the issues of the AGC in two-area interconnected power systems by proposing a new approach based on both Jaya algorithm and the rank exponent method. In particular, we design a proportional-integral-derivative controller with derivative filtering (PIDm), where the effect of the noise is mitigated by the use of a filter with derivative gain. We propose to build the objective function, to tune the controller's parameters, as the linear combination of three sub-objectives, namely integral of time multiplied absolute error (ITAE) for frequency deviations, tie-line power deviation, and area-control errors (ACEs). The rank method is exploited to evaluate the weights of these sub-objectives, while the final overall objective function is minimized exploiting the Jaya algorithm. The proposed controller's performance is assessed in six different scenarios with load disturbances, and its effectiveness is compared to state-of-art controllers tuned using salp swarm algorithm (SSA), Nelder-Mead simplex (NMS), symbiotic organisms search (SOS), elephant herding optimization (EHO), and Luus-Jaakola (LJ) optimization algorithms. To illustrate the frequency and tie-line power changes, results are also shown, and a statistical study is finally carried out to evaluate the recommended controller's overall effectiveness. Additionally, Friedman rank test as no-parametric statistical analysis is also done in order to evaluate the significance level of optimization algorithms. Our numerical findings evidence that the proposed PIDm controller outperforms other existing optimization-based controllers in terms of performance and utility, thus proving to be very effective for handling AGC issues in two-are interconnected power systems.
computer science, information systems,telecommunications,engineering, electrical & electronic