Magnetic Pose Estimation Using Distributed Dipole Models

This project studies magnetic pose estimation for flexible permanent magnets using distributed dipole models and magnetic sensor arrays.

The research focuses on solving nonlinear inverse problems arising from magnetic field measurements collected by a dense array of sensors. Instead of approximating the magnet as a single dipole, the system models the magnet as multiple distributed dipoles, enabling accurate tracking of bending and deformation.

Key components of the project include:

• Modeling flexible magnets using distributed dipole representations
• Real-time magnetic field sensing with multi-sensor arrays
• Nonlinear least-squares optimization for pose estimation
• Development of Python-based algorithms for sensor calibration and data processing
• Visualization and reconstruction of magnet shape and position

The work combines physics-based modeling with computational optimization to enable accurate tracking of soft magnetic structures for applications in soft robotics, tactile sensing, and intelligent materials.