Paired Interactions of Magnetic Millirobots in Confined Spaces Through Data-Driven Disturbance Rejection Control Under Global Input

Owing to their high task efficiency and load capacity in closed space operations, multiple millirobots system has drawn extensive attention recently. However, the limited global magnetic fields and nonlinear interactions between individual robots make it challenging to control multiple millirobots in close proximity to each other, resulting in difficulty in achieving accurate paired interactions. Here, we propose a paired interactive control method for multiple millirobots, which enables the precise formation of two millirobots within a multiple millirobot system. The paired interactive motion is modeled within a singular point tracking framework to facilitate the implementation of an independent control strategy for multiple microrobots. Then, a data-driven actuation-movement mapping model for two millirobots is established as a nonlinear inversion controller, enabling the control system to rapidly achieve the desired state. To eliminate residual errors, a feedback controller is designed based on the active disturbance rejection concept, which estimates and eliminates generalized disturbances via an extended state observer. The control method is validated by accurately achieving planar formations via two millirobots both in isolation and within a multiple millirobot system, in which the root mean square error is less than 3% of the single-robot length.