Analytical Target Cascading Based Co-Optimization of Transmission and Distribution Systems With Energy Storage System

The integration of renewable energy sources and energy storage systems (ESS) in transmission and distribution networks poses significant challenges in optimizing the power flow and scheduling the ESS efficiently which requires addressing intricate time-domain coupling constraints. To address this issue, we propose a novel Bi-level co-optimization framework called the Q-learning based Analytical Target Cascading (ATC-Q) optimization for solving alternating current optimal power flow (ACOPF) in transmission and distribution combined networks. The algorithm considers ESS at both T&D systems over a 24-hour receding horizon, taking into account real-time energy prices, current ESS state of charge, forecasted photovoltaics (PV), and load demand. The proposed solution aims to improve the scheduling of ESS by decoupling time-domain constraints and formulating a queue based on current and future ESS states. The solution is seamlessly integrated into the iterative process of ATC, which considers two-stage stochastic Transmission and Distribution AC Optimal Power Flow (T&D ACOPF) into a upper level and sub level problem. The computational efficiency is maintained by sharing relevant information between each level. To validate the proposed approach, the algorithm is implemented on a modified IEEE 39 Bus transmission grid network connected to a 34 Bus distribution network, incorporating photovoltaics, ESS, and dynamic load. Through comprehensive case studies, the proposed algorithm showcased a reduction in operational cost and enhanced combined network operation. The proposed algorithm is validated using a Controller Hardware-in-the-Loop setup implemented using a digital real-time simulator (DRTS) and an external computer serving as the main controller, showcasing its cost-effectiveness and suitability for real-time applications.