A Brain Inspired Learning Algorithm for the Perception of a Quadrotor in Wind

The quest for a brain-inspired learning algorithm for robots has culminated in the free energy principle from neuroscience that models the brain's perception and action as an optimization over its free energy objectives. Based on this idea, we propose an estimation algorithm for accurate output prediction of a quadrotor flying under unmodelled wind conditions. The key idea behind this work is the handling of unmodelled wind dynamics and the model's non-linearity errors as coloured noise in the system, and leveraging it for accurate output predictions. This paper provides the first experimental validation for the usefulness of generalized coordinates for robot perception using Dynamic Expectation Maximization (DEM). Through real flight experiments, we show that the estimator outperforms classical estimators with the least error in output predictions. Based on the experimental results, we extend the DEM algorithm for model order selection for complete black box identification. With this paper, we provide the first experimental validation of DEM applied to robot learning.

[1]  Ajith Anil Meera,et al.  Dynamic Expectation Maximization Algorithm for Estimation of Linear Systems with Colored Noise , 2021, Entropy.

[2]  Martijn Wisse,et al.  Free Energy Principle for State and Input Estimation of a Quadcopter Flying in Wind , 2021, 2022 International Conference on Robotics and Automation (ICRA).

[3]  Tim Verbelen,et al.  Robot navigation as hierarchical active inference , 2021, Neural Networks.

[4]  Thomas Parr,et al.  Generative Models for Active Vision , 2021, Frontiers in Neurorobotics.

[5]  Martijn Wisse,et al.  Free Energy Principle Based State and Input Observer Design for Linear Systems with Colored Noise , 2020, 2020 American Control Conference (ACC).

[6]  Riccardo M. G. Ferrari,et al.  A Novel Adaptive Controller for Robot Manipulators Based on Active Inference , 2019, IEEE Robotics and Automation Letters.

[7]  Gordon Cheng,et al.  Active inference body perception and action for humanoid robots , 2019, ArXiv.

[8]  Manuel Baltieri,et al.  PID Control as a Process of Active Inference with Linear Generative Models † , 2019, Entropy.

[9]  Simon McGregor,et al.  The free energy principle for action and perception: A mathematical review , 2017, 1705.09156.

[10]  Yong Zhang,et al.  Unbiased identification of a class of multi-input single-output systems with correlated disturbances using bias compensation methods , 2011, Math. Comput. Model..

[11]  Yong Zhang,et al.  Bias compensation methods for stochastic systems with colored noise , 2011 .

[12]  Karl J. Friston,et al.  Action understanding and active inference , 2011, Biological Cybernetics.

[13]  Karl J. Friston,et al.  Action and behavior: a free-energy formulation , 2010, Biological Cybernetics.

[14]  Karl J. Friston,et al.  The default-mode, ego-functions and free-energy: a neurobiological account of Freudian ideas , 2010, Brain : a journal of neurology.

[15]  Karl J. Friston The free-energy principle: a unified brain theory? , 2010, Nature Reviews Neuroscience.

[16]  Karl J. Friston,et al.  Reinforcement Learning or Active Inference? , 2009, PloS one.

[17]  Karl J. Friston The free-energy principle: a rough guide to the brain? , 2009, Trends in Cognitive Sciences.

[18]  Karl J. Friston,et al.  Predictive coding under the free-energy principle , 2009, Philosophical Transactions of the Royal Society B: Biological Sciences.

[19]  Karl J. Friston Hierarchical Models in the Brain , 2008, PLoS Comput. Biol..

[20]  Karl J. Friston,et al.  DEM: A variational treatment of dynamic systems , 2008, NeuroImage.

[21]  Wei Xing Zheng,et al.  On a least-squares-based algorithm for identification of stochastic linear systems , 1998, IEEE Trans. Signal Process..

[22]  M. Wisse,et al.  On the Convergence of DEM's Linear Parameter Estimator , 2021, PKDD/ECML Workshops.

[23]  D. Benders AR.Drone 2.0 state estimation using Dynamic Expectation Maximization: Bringing brain perception theory to practice , 2020 .

[24]  Enrique Alarcón Álvarez,et al.  Using the EM algorithm to estimate the state space modelfor OMAX , 2014 .

[25]  Lennart Ljung,et al.  System Identi cation , 2014 .

[26]  Jing Lu,et al.  Least squares based iterative identification for a class of multirate systems , 2010, Autom..

[27]  Biao Huang,et al.  System Identification , 2000, Control Theory for Physicists.