Design and Real-Time Optimization for a Magnetic Actuation System With Enhanced Flexibility

Abstract

In this article, a magnetic actuation system based on three mobile electromagnetic coils is designed and a control strategy for the system is proposed. Enhanced flexibility combined with optimization algorithms enables the system to satisfy various requirements in applications, such as avoiding collision between the coils and the obstructions within the workspace, placing the coils to optimal positions to enhance energy efficiency, generating wide varieties of magnetic field for complex tasks, and tracking the location of the robot with enlarged workspace. To reach that purpose, a model of the system is built for magnetic field calculation, and a real-time optimization algorithm based on particle swarm optimization combined with a collision detection algorithm is proposed and implemented to calculate optimal positions for coils and at the same time avoid collision. We fabricate a prototype system, named RoboMag, to prove the concept. Simulations and experiments on helical swimmer and soft robot are conducted to evaluate the performance. Compared with two conventional control strategies, the demanded currents for long-distance actuation are reduced by up to 62.7%. The calculation process is conducted in real time and the coils are able to avoid collision with the barriers inside the workspace during actuation. Moreover, generation and steering of a microrobotic swarm is demonstrated, showing the capability of the system in generating programmed dynamic fields for complicated tasks.