A MILP Model Incorporated With the Risk Management Tool for Self-Healing Oriented Service Restoration

The inevitable emergence of intelligent distribution networks has introduced new features in these networks. According to most experts, self-healing is one of the main abilities of smart distribution networks. This feature increases the reliability and resiliency of networks by reacting fast and restoring the critical loads (CLs) during a fault. Nevertheless, the stochastic nature of the components in a power system imposes significant computational risk in enabling the system to self-heal. In this paper, a mathematical model is introduced for the self-healing operation of networked Microgrids (MGs) to assess the risk in the optimal service restoration (SR) problem. Electric vehicles (EVs) and plug-in hybrid electric vehicles (PHEVs) and their stochastic nature besides the distributed generation units (DGs), the ability to reconfiguration, and demand response program are considered simultaneously. The objective function is designed to maximize the restored loads and minimize the risk. The Conditional Value-at-Risk (CVaR) is used to calculate the risk of the SR as one of the most efficient and famous risk indices. In the general case study and considering β equal to the 0, 1, 2, 3, and 4, expected values of SR for the risk-averse problem is 21.2, 20, 19.3, 19.1, and 19% less than the risk-neutral problem, respectively. The formulation of the problem is mixed-integer linear programming (MILP), and the model is tested in the modified Civanlar test system. The analysis of several case studies has proved the performance of the proposed model and the importance of risk management in the problem.