With the widespread adoption of bidirectional converters, modern metro traction power supply systems (TPSS) are becoming increasingly flexible and controllable. Rapid advancements in rail transit technologies demand interdisciplinary talents beyond traditional single-discipline electrical engineering expertise to address the complexities of these systems. Conventional teaching methods, often limited by disciplinary silos and inadequate hands-on experimentation, fall short in fostering students’ practical engineering cognition. In response to these structural limitations, this study develops an experimental simulation platform for load flow calculations in flexible TPSS.

Conventional iterative load flow calculation methods, such as the Newton–Raphson approach, suffer from the convergence and initialization limitations when applied to flexible TPSS. Accordingly, the holomorphic embedding load-flow method (HELM) is adopted as a noniterative deterministic algorithm for solving nonlinear load flow equations. HELM offers excellent global convergence, independence from initial value guesses, and elimination of Jacobian matrix computations, making it particularly suitable for this application. First, an equivalent mathematical model of the flexible TPSS is constructed. Based on this model, a unified load flow framework is established using the principles of holomorphic function embedding, encompassing AC and DC systems as well as bidirectional converters. Second, recursive relationships are derived between coefficients in the low- and high-order power series expansions of the holomorphic functions, enabling analytical solutions to the load flow model. Furthermore, to minimize complex matrix operations and enhance computational efficiency, the Padé approximation technique based on Bauer’s Eta algorithm is adopted to obtain analytical solutions for the target voltage functions.

Experimental results yield the following insights: 1) A timed simulation model of a three-station, two-train system validates the accuracy and feasibility of the proposed platform for load flow calculations in TPSS. 2) Under a power control strategy incorporating bidirectional converters, the system effectively mitigates excessive traction power demands, maintaining substation power within the safe operating limits and markedly enhancing overall system stability. 3) Implementing droop control enables automatic power sharing among converters; a larger droop coefficient improves load distribution balance but amplifies voltage fluctuations. Therefore, careful selection of the droop coefficient is essential to achieve an optimal trade-off between power sharing and voltage stability.

This study designs an experimental simulation platform for load flow calculations in flexible TPSS for urban rail transit. The platform assists students in conducting analysis and learning activities related to these systems, including system modeling, HELM application, and tuning of bidirectional converter control strategies. It enhances students’ comprehension of flexible TPSS, bridges theoretical knowledge with engineering practice, and lays a strong foundation for training high-level professionals needed for the development of intelligent, low-carbon urban rail transit TPSS.