Machine Learning-Enabled Precision Position Control and Thermal Regulation in Advanced Thermal Actuators
This work addresses a practical limitation in robotics by enabling sensorless control of high-performance artificial muscles, though it appears incremental as it builds on existing methods for thermal actuators.
The researchers tackled the problem of controlling nylon artificial muscles without external sensors by developing a machine learning-based constant power open-loop controller, achieving precise position control and thermal regulation.
With their unique combination of characteristics - an energy density almost 100 times that of human muscle, and a power density of 5.3 kW/kg, similar to a jet engine's output - Nylon artificial muscles stand out as particularly apt for robotics applications. However, the necessity of integrating sensors and controllers poses a limitation to their practical usage. Here we report a constant power open-loop controller based on machine learning. We show that we can control the position of a nylon artificial muscle without external sensors. To this end, we construct a mapping from a desired displacement trajectory to a required power using an ensemble encoder-style feed-forward neural network. The neural controller is carefully trained on a physics-based denoised dataset and can be fine-tuned to accommodate various types of thermal artificial muscles, irrespective of the presence or absence of hysteresis.