Miaozhang Shen

Physics-Informed Residual Network for Magnetic Dipole Model Correction and High-Accuracy Localization

The magnetic dipole model exhibits significant deviations from real-world sensor data due to neglected material nonlinearities and environmental interference. This paper proposed a Physics-Informed Residual Network (PIRNet) that adaptively corrected simulated magnetic field data by integrating dipol