PhD researcher: Increasing the Hosting Capacity of Distribution System for Electric Vehicles with Active Network Management

With the high penetration of renewable energy resources (RES) in the power system, due to the variety and intermittency that characterize most of them, the degree of uncertainty in both the generation and demand side of the power system has increased. So operators and planners should consider the risks that uncertainty brings, to deal with its consequences (i.e., more flexibility is needed to operate a power system in the safe and stable zone). Demand response is one of the main approaches to provide sufficient flexibility. On the one hand, the shift to a more active, decentralized, and complex system creates a task that can be more unmanageable for humans. On the other hand, it is essential to consider privacy for residential resources in a flexible demand response. So the advent of new technologies such as cyber-security and artificial intelligence creates an opportunity to alleviate the load on humans by assisting and partially automating to address these problems. By utilizing these approaches, at first, the residential resources data are kept locally and privately, and then the uncertainties brought by the flexible resources can be captured. Setting more flexible resources (such as residential resources) in the power system makes the operation more complicated. So, to operate the system more secure and less costly, the optimization methods can be used by considering the security constraints. However, in some cases that the flexibility requirement of the system should be predicted (i.e., prediction of charging energy consumption of individual EV with a specific time resolution), machine learning and deep learning methods can work out to make more accurate predictions.