Residual Physics-Informed Neural Networks for High-Fidelity BLDC Motor Modeling
For robotic joint control, this provides a fast, accurate surrogate model for real-time motor state estimation and control, but the approach is incremental as it applies existing PINN and ResNet techniques to a specific domain.
This paper introduces a residual physics-informed neural network (PINN) that models BLDC motor dynamics, achieving up to 118x faster inference than conventional ODE solvers (0.1–22 μs per query) while maintaining accuracy. Training completes in under two minutes on a CPU.
Accurate dynamics modeling of Brushless DC (BLDC) motors is fundamental to high-performance robotic joint control. This paper presents a Physics-Informed Neural Network (PINN) with a deep residual (ResNet) backbone that learns a continuous-time surrogate of the full six-state BLDC motor dynamics. Given simulation time, applied three-phase voltages, and excitation parameters as inputs, the network directly predicts all motor state variables -- rotor angle, angular velocity, three-phase currents, and winding temperature -- while simultaneously satisfying the governing electromechanical and thermal ODEs through a composite physics-data loss. A curriculum scheduling strategy gradually activates the physics penalty to prevent premature convergence. Training runs are completed in under two minutes on a standard CPU. Crucially, once trained, PINN inference achieves latencies of 0.1--22, mu s per query, up to 118x faster than conventional ODE solvers, making it suitable for real-time observer and control applications.