Abstract: The parameters of the synchronous switched reluctance motor (SSRM) vary with operating conditions and load variations, which affects the control performance of the motor. To address this issue, an online parameter identification method based on a physics-informed neural network (PINN) was proposed for the identification of stator resistance and d-q axis inductances. The proposed method combined data-driven learning with physics-based modeling. A loss function was constructed using both the mean squared error of the identified parameters and the voltage equations of the SSRM, enabling the neural network to be jointly constrained by measurement data and physical laws during training. Simulation and experimental results show that the proposed method achieves high identification accuracy. The identification error of stator resistance is 7.16%, the identification error of d-axis inductance is 1.52%, and the identification error of q-axis inductance is 2.4%. Compared with the RLSFF algorithm, this algorithm has a faster convergence speed and can effectively reduce the identification errors of stator resistance and stator inductance.