A Positioning System of Capsule Endoscope Based on the Combination of Particle Swarm Optimization and L-M Algorithm

In this paper, we design a positioning system for permanent magnet based on three-axis digital magnetic sensor array to achieve accurate positioning of capsule endoscope. The permanent magnet is equivalent to a magnetic dipole model and fixed in the capsule, the host computer receives the magnetic flux density data measured by the magnetic sensor array through serial port. We propose a method combining particle swarm optimization and Levenberg-Marquard (L-M) algorithm to calculate position and direction parameters of the permanent magnet, and display real-time trajectory tracking through the graphical interface in MATLAB. The system takes 0.2 seconds to complete positioning once, and positioning error of the system is 0.479 cm.