Availability augmentation of smart grid systems using Markovian approach and nature inspired algorithms

The present study focuses on the optimization of availability in smart grid systems by integrating stochastic modeling and hybrid nature-inspired algorithms. For this purpose, a novel stochastic model of smart grid systems, comprising seven subsystems under various redundancy strategies, is developed using Markov birth-death process. The concept of degradation, uncertain sensor behavior and exponentially distributed failure and repair laws utilized for system performance evaluation. Nature-inspired algorithms employed for parameter estimation and availability optimization at various population sizes. The efficiency of various algorithms for smart grid system availability prediction is statistically tested using nonparametric statistical method, namely Friedman test, followed by post hoc Wilcoxon test. The optimizer results are compared with the simulated system availability results to verify the accuracy and robustness of the proposed model. It is revealed that sequential SOHBA outperforms the traditional SO and HBA algorithms as well as Monte Carlo simulation results. The numerical results of steady-state availability are also derived and impact of variation in failure and repair rates investigated. To highlight the importance of study numerical and graphical results are appended.