Active Inference for Integrated State-Estimation, Control, and Learning

This work presents an approach for control, state-estimation and learning model (hyper)parameters for robotic manipulators. It is based on the active inference framework, prominent in computational neuroscience as a theory of the brain, where behaviour arises from minimizing variational free-energy. First, we show there is a direct relationship between active inference controllers, and classic methods such as PID control. We demonstrate its application for adaptive and robust behaviour of a robotic manipulator that rivals state-of-the-art. Additionally, we show that by learning specific hyperparameters, our approach can deal with unmodeled dynamics, damps oscillations, and is robust against poor initial parameters. The approach is validated on the ‘Franka Emika Panda’ 7 DoF manipulator. Finally, we highlight limitations of active inference controllers for robotic systems.

Authors

• Mohamed Baioumy
• Paul Duckworth
• Bruno Lacerda
• Nick Hawes

Keywords

• PI control
• Conferences
• Computational neuroscience
• Estimation
• Robustness
• Filtering theory
• PD control
• Active Inference
• Hyperparameters
• Adaptive Behavior
• Proportional-integral-derivative
• Robot Manipulator
• Free Energy
• Dynamical
• Covariance Matrix
• Gradient Descent
• Optimal Control
• Tuning Parameter
• Kullback-Leibler
• Adaptive Control
• True State
• Explicit Model
• Hidden State
• Target State
• Estimation Step
• Temporal Parameters
• Belief State
• Factor Graph
• State Transition Model
• Positional Encoding
• Evidence Lower Bound
• Joint Velocity
• Error Term