The miniature-sized magnetic microrobot capable of swimming in liquid with a low Reynolds number shows promise in targeted therapy because it can move flexibly in a narrow environment.
However, due to the impacts of non-linearity and the diversity of the complex trajectories desired, it is difficult to ensure the accuracy of microrobot tracking without frequent controller adjustment.
Recently, a research team led by XU Sheng and XU Tiantian from the Shenzhen Institute of Advanced Technology of the Chinese Academy of Sciences developed a new method based on the Wide Learning System (BLS), which can achieve a precise and flexible path tracking of the microrobot.
The study was published in IEEE Cybernetics Transactions November 1.
Compared with the traditional deep learning method, BLS with simple and flexible structure could achieve satisfactory accuracy.
The research team developed a learning-based microrobot servo control algorithm using BLS and combined Lyapunov theory with the complex learning method to derive the constraints of the controller parameters.
They also developed a BLS-based controller training algorithm that uses several desired tracking paths as demonstrations and obtained the controller parameters by the training algorithm.
According to the simulation and experimental results, this BLS-based method had a fast training speed, which only took about 6 seconds.
The trained controller based on the BLS can follow the trajectories with different shapes and speeds with better precision, and it does not require any readjustment of the parameters due to its strong generalization ability. In addition, since the BLS method is applied, the number of nodes can be flexibly adjusted when further demonstrations are needed.