Cascades-Risk Informed Transmission Expansion Planning by Behavioural Cloning

The growing energy demand, coupled with the increasing trend towards electrification, challenges Transmission Expansion Planning (TEP). In this work, we embrace a risk-informed approach to TEP and address it as a multi-objective optimization problem considering the system risks and the expansion costs. Genetic algorithms, such as NSGAII, are effective in tackling multi-objective optimization problems. We here explore the potential of reinforcement learning as an alternative methodology. Specifically, we used imitation learning techniques to gain insight from prior solutions generated using NSGAII. This novel approach lays the groundwork for future research on reinforcement learning methods, based on policies acquired through imitation learning techniques. By adopting this innovative framework, we anticipate significant advancements in the efficiency and effectiveness of TEP decision-making processes.