How Can a Simulator Replicate Real-World EV Charging?

Electric cars are reshaping how we use energy, yet their charging still puzzles power grids. To get a clearer picture, researchers built an EV charger model on Typhoon HIL, a real‑time circuit simulation platform. The model reproduces current dynamics, control logic, and even Battery Management System (BMS) decisions, such as stopping the charge at 97% state‑of‑charge to protect the battery. Simulation results were compared with real data collected by phasor measurement units (PMUs) installed at a charging station. The mean absolute error (MAE) was 0.14 kW and the root‑mean‑square error (RMSE) 0.36 kW, showing an almost perfect match. Thanks to this digital twin, smart‑charging strategies can be tested, grid impacts assessed, and infrastructure optimized without real‑world risks. In practice, the simulator becomes a virtual lab where engineers and researchers experiment with new solutions for greener mobility and a more stable grid.

The physics‑based model implemented on Typhoon HIL relies on differential equations describing the battery V‑I relationship, incorporating internal resistance, electrochemical capacity, and thermal effects. Simulation runs with a 100 µs time step, providing the resolution needed to capture current transients during pre‑charge, constant‑current, and constant‑voltage phases. The BMS is programmed to monitor SOC using a coulomb‑counting algorithm, adding a voltage‑based correction to reduce estimation error. The 97% SOC termination strategy mimics the common practice of preserving cell life by avoiding over‑charging. Validation data come from PMUs sampling at 50 Hz, synchronized via GPS to ensure temporal alignment with the simulation. Statistical analysis yielded an MAE of 0.14 kW and an RMSE of 0.36 kW, both below the typical 0.5 kW acceptance threshold for charging models. Moreover, the framework supports integration of model predictive control (MPC) and machine‑learning algorithms, allowing testing of flexible charging scenarios in response to price signals or grid capacity limits. This approach paves the way for full digital twins capable of real‑time interaction with energy management systems (EMS) and providing reliable simulation data for long‑term planning studies.