Fuel Cell Low-Floor Tram Power System Cable Crosstalk Prediction Model

The fuel cell system encompasses numerous electrical components tightly integrated into one entity. In terms of internal layout and interface wiring, it faces a complex and variable electromagnetic environment. Cables play a crucial role in the power system, and their radiation and immunity performance are essential for the overall electromagnetic compatibility of the system. Studying the crosstalk coupling issues between cables is particularly important. Traditional electromagnetic calculation methods lack wide applicability and self-learning potential. Therefore, we adopted the FFA to optimize the support vector regression model for establishing a crosstalk model between cables and predicting crosstalk voltage. We successfully established mapping relationships between various factors and cable crosstalk, thereby achieving accurate predictions of crosstalk voltage, The RMSE metric is only 0.0585, and the $R^{2}$ score reaches 0.9711. This approach not only contributes to a deeper understanding of the crosstalk mechanism but also provides a reliable predictive framework for future issues.