Probabilistic modeling and optimization of microgrids with EV parking lots and dispersed generation

Abstract

Distributed generation sources provide self-governing power during outages, making microgrids and islanded distribution networks vital for service endurance, superior power quality, reliability, and operative efficiency. However, microgrids structure are difficult to control, particularly in islanded mode where no main power source exists if the main grid fails. Fast responses from discrete generation sources using power electronics can undermine the grid during faults or normal operations without proper regulations. The double-fed induction generator (DFIG) has become the preferred wind turbine generator owing to its low cost and flexibility to varying wind speeds. This paper presents a probabilistic scheduling for day-ahead microgrid programming that includes EV parking lots and dispersed generation resources. The microgrid works in both normal and islanded modes depending on main grid conditions. The uncertainty in EV parking lot usage is modeled hourly using the Z-number method, while wind and solar generation, market prices, and loads are modeled using the Monte Carlo method. Scenario-based incidents in the upstream grid that lead to microgrid islanding are considered, focusing on the time and duration of impact. The optimization model accounts for uncertainty, EV charging/discharging, and operational costs under normal and fault conditions. The fault ride-through (FRT) method for maintaining DFIG stability in islanded microgrids are proposed. In this technique stabilizes terminal voltage during faults by employing a resistor in series with the DFIG stator, enhancing voltage stability and FRT capability. Without these methods, the DFIG may lose stability after clearing transient errors, risking generator loss and threatening microgrid stability, particularly in islanded mode. The effectiveness of these control and protection strategies is validated through comprehensive simulations in MATLAB.