Optimizing Configuration of Cyber Network Considering Graph Theory Structure and Teaching–Learning-Based Optimization (GT-TLBO)

Selecting the best configuration for a cyber-network system is considered as one of the major challenges in terms of the reliability evolution of smart systems because of a specific relationship between power network system and cyber network. Although producing some possible cyber networks for one simple power system and selecting the best one does not need any formulas and can be done through trial and error as performed in previous studies, producing all possible nth elements of the cyber network that observe the cyber protocols and choosing the best one has not been done before. Choosing the best configuration for any nth elements cyber network that can be connected to a bulk power network needs a robust and adequate computer algorithm considering cyber protocols and decision-making goals. In this paper, a novel method is proposed in order to introduce the best configuration for a cyber-network system to accommodate cyber protocols and have a remarkable effect on decreasing expected energy not supplied (EENS) compared with those previously studied. To this end, two mathematical concepts are proposed; a graph theory to consider cyber protocols and teaching-learning-based optimization to select the best cyber configuration with minimum EENS. During the first time, choosing the best configuration with each specific goal for nth devices of a cyber-network system is applicable by converting it into an n × n adjacency matrix and using the proposed graph theory mixed with teaching-learning-based optimization algorithm. Moreover, Monte Carlo simulation was used in this paper as one of the most precise probabilistic methods. The test results indicate that the proposed method selected for identifying the best configuration of a cyber-network system has significantly improved reliability indices compared to those in previous papers and it could be useful for every wide power-cyber network. Additionally, different types of cyber network are studied for validating the proposed method. This method is applied to the realistic feeder of the Hormozgan Regional Electric Company as a smart pilot system.