Fault-tolerant control of robot manipulators with sensory faults using unbiased active inference

This work presents a novel fault-tolerant control scheme based on active inference. Specifically, a new formulation of active inference which, unlike previous solutions, provides unbiased state estimation and simplifies the definition of probabilistically robust thresholds for fault-tolerant control...

Full description

Bibliographic Details
Main Authors: Baioumy, M, Pezzato, C, Ferrari, R, Corbato, CH, Hawes, N
Format: Conference item
Language:English
Published: IEEE 2022
_version_ 1797087399896940544
author Baioumy, M
Pezzato, C
Ferrari, R
Corbato, CH
Hawes, N
author_facet Baioumy, M
Pezzato, C
Ferrari, R
Corbato, CH
Hawes, N
author_sort Baioumy, M
collection OXFORD
description This work presents a novel fault-tolerant control scheme based on active inference. Specifically, a new formulation of active inference which, unlike previous solutions, provides unbiased state estimation and simplifies the definition of probabilistically robust thresholds for fault-tolerant control of robotic systems using the free-energy. The proposed solution makes use of the sensory prediction errors in the free-energy for the generation of residuals and thresholds for fault detection and isolation of sensory faults, and it does not require additional controllers for fault recovery. Results validating the benefits in a simulated 2-DOF manipulator are presented, and future directions to improve the current fault recovery approach are discussed.
first_indexed 2024-03-07T02:35:11Z
format Conference item
id oxford-uuid:a889578a-afa0-481c-9259-136ad1343a5c
institution University of Oxford
language English
last_indexed 2024-03-07T02:35:11Z
publishDate 2022
publisher IEEE
record_format dspace
spelling oxford-uuid:a889578a-afa0-481c-9259-136ad1343a5c2022-03-27T03:02:20ZFault-tolerant control of robot manipulators with sensory faults using unbiased active inferenceConference itemhttp://purl.org/coar/resource_type/c_5794uuid:a889578a-afa0-481c-9259-136ad1343a5cEnglishSymplectic ElementsIEEE2022Baioumy, MPezzato, CFerrari, RCorbato, CHHawes, NThis work presents a novel fault-tolerant control scheme based on active inference. Specifically, a new formulation of active inference which, unlike previous solutions, provides unbiased state estimation and simplifies the definition of probabilistically robust thresholds for fault-tolerant control of robotic systems using the free-energy. The proposed solution makes use of the sensory prediction errors in the free-energy for the generation of residuals and thresholds for fault detection and isolation of sensory faults, and it does not require additional controllers for fault recovery. Results validating the benefits in a simulated 2-DOF manipulator are presented, and future directions to improve the current fault recovery approach are discussed.
spellingShingle Baioumy, M
Pezzato, C
Ferrari, R
Corbato, CH
Hawes, N
Fault-tolerant control of robot manipulators with sensory faults using unbiased active inference
title Fault-tolerant control of robot manipulators with sensory faults using unbiased active inference
title_full Fault-tolerant control of robot manipulators with sensory faults using unbiased active inference
title_fullStr Fault-tolerant control of robot manipulators with sensory faults using unbiased active inference
title_full_unstemmed Fault-tolerant control of robot manipulators with sensory faults using unbiased active inference
title_short Fault-tolerant control of robot manipulators with sensory faults using unbiased active inference
title_sort fault tolerant control of robot manipulators with sensory faults using unbiased active inference
work_keys_str_mv AT baioumym faulttolerantcontrolofrobotmanipulatorswithsensoryfaultsusingunbiasedactiveinference
AT pezzatoc faulttolerantcontrolofrobotmanipulatorswithsensoryfaultsusingunbiasedactiveinference
AT ferrarir faulttolerantcontrolofrobotmanipulatorswithsensoryfaultsusingunbiasedactiveinference
AT corbatoch faulttolerantcontrolofrobotmanipulatorswithsensoryfaultsusingunbiasedactiveinference
AT hawesn faulttolerantcontrolofrobotmanipulatorswithsensoryfaultsusingunbiasedactiveinference