NN FMU: Deep Learning Models for Next-Gen xiL Vehicle Simulation

Functional Mock-up Units (FMUs) have become a standard for enabling co-simulation and model exchange in vehicle system development. However, traditional FMUs derived from physics-based models can be computationally intensive, especially in scenarios requiring real-time performance. This paper presents a Python-based approach for developing a Neural Network (NN) based FMU using deep learning techniques, aimed at accelerating vehicle simulation while ensuring high fidelity. The neural network was trained on vehicle simulation data and developed using Python frameworks such as TensorFlow. The trained model was then exported into FMU, enabling seamless integration with FMI-compliant platforms. The NN FMU replicates thermal behavior of vehicle with high accuracy while offering a significant reduction in computational load. Benchmark comparisons with Physical Thermal Model demonstrate that the proposed solution provides both efficiency and reliability across various driving conditions. Additionally, the paper discusses the workflow for model training and integration strategies for deep learning models within source simulation tools like AMESIM and Simulink. This paper covers methodology of converting Thermal Physical systems of vehicle into NN FMU and replace the physical sub-system by NN FMU. By doing this significant time reduction is observed without affecting the accuracy when compared with Physical sub-system model. NN FMU also reduces efforts up to 40 % compared with traditional FMU conversion. Also, CPU improvement from Physical to NN FMU model achieved greater than 20 % Reduction with same accuracy. NN FMU maintains FMI compatibility and can be directly used in a wide range of XiL applications such as Model-in-the-Loop (MiL), Software-in-Loop (SiL), and Hardware-in-Loop (HiL) testing scenarios. This NN FMUs opens pathways for hybrid modelling approaches that combine data-driven and physics-based paradigms for automotive simulations