Research on Real-Time Operation Optimization of the Urban Rail Traction Power Supply System Integrated Storage and Feedback

Urban rail transit plays an important role in the rapid economic and social development of China. The integrated storage and feedback urban rail traction power supply system is one of the strategies for the green and low-carbon development of urban rail transit. This paper establishes a real-time operation optimization model based on the Markov decision process in the context of the integrated storage and feedback urban rail traction power supply system. It uses offline training of a deep reinforcement learning agent to optimize the control parameters of the energy feedback systems and energy storage systems in real-time to reduce the traction energy consumption of the power supply system, providing a reference for energy-saving operation of subways.