Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AI-Enhanced Kinematic Modeling of Flexible Manipulators Using Multi-IMU Sensor Fusion

Amir Hossein Barjini, Jouni Mattila

TL;DR

This work tackles the challenge of accurately estimating the pose of flexible manipulators using low-cost multi-IMU sensors under gravity and vibrations. It integrates a chain-of-rigid-segments kinematic model with a complementary filter whose gains are optimized by PSO, and further refines residual errors through an RBFNN trained on ground-truth data. The approach demonstrates marked accuracy improvements, achieving RMSEs on the order of 2e-4 m for $y$, 4e-4 m for $z$, and 2.4e-4$ rad for $\theta$, verified on a 4.5 m steel link with a distributed IMU network and laser-tracker ground truth. The framework enables accurate, cost-effective end-effector state estimation suitable for industrial deployment and adaptable to rigid, flexible, and soft robotic systems.

Abstract

This paper presents a novel framework for estimating the position and orientation of flexible manipulators undergoing vertical motion using multiple inertial measurement units (IMUs), optimized and calibrated with ground truth data. The flexible links are modeled as a series of rigid segments, with joint angles estimated from accelerometer and gyroscope measurements acquired by cost-effective IMUs. A complementary filter is employed to fuse the measurements, with its parameters optimized through particle swarm optimization (PSO) to mitigate noise and delay. To further improve estimation accuracy, residual errors in position and orientation are compensated using radial basis function neural networks (RBFNN). Experimental results validate the effectiveness of the proposed intelligent multi-IMU kinematic estimation method, achieving root mean square errors (RMSE) of 0.00021~m, 0.00041~m, and 0.00024~rad for $y$, $z$, and $θ$, respectively.

AI-Enhanced Kinematic Modeling of Flexible Manipulators Using Multi-IMU Sensor Fusion

TL;DR

This work tackles the challenge of accurately estimating the pose of flexible manipulators using low-cost multi-IMU sensors under gravity and vibrations. It integrates a chain-of-rigid-segments kinematic model with a complementary filter whose gains are optimized by PSO, and further refines residual errors through an RBFNN trained on ground-truth data. The approach demonstrates marked accuracy improvements, achieving RMSEs on the order of 2e-4 m for , 4e-4 m for , and 2.4e-4\theta$, verified on a 4.5 m steel link with a distributed IMU network and laser-tracker ground truth. The framework enables accurate, cost-effective end-effector state estimation suitable for industrial deployment and adaptable to rigid, flexible, and soft robotic systems.

Abstract

This paper presents a novel framework for estimating the position and orientation of flexible manipulators undergoing vertical motion using multiple inertial measurement units (IMUs), optimized and calibrated with ground truth data. The flexible links are modeled as a series of rigid segments, with joint angles estimated from accelerometer and gyroscope measurements acquired by cost-effective IMUs. A complementary filter is employed to fuse the measurements, with its parameters optimized through particle swarm optimization (PSO) to mitigate noise and delay. To further improve estimation accuracy, residual errors in position and orientation are compensated using radial basis function neural networks (RBFNN). Experimental results validate the effectiveness of the proposed intelligent multi-IMU kinematic estimation method, achieving root mean square errors (RMSE) of 0.00021~m, 0.00041~m, and 0.00024~rad for , , and , respectively.

Paper Structure

This paper contains 24 sections, 42 equations, 7 figures, 1 table.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Motivation and Research Gap
  4. Research Contributions and Structure of the Paper
  5. Kinematic Modeling Using IMU-Based Sensor Fusion
  6. Modeling the Flexible Manipulator as a Chain of Rigid segments
  7. IMU-Based Orientation and Motion Derivation
  8. Gravity-Referenced Joint Angle Estimation
  9. Joint Angular Velocity Estimation
  10. Complementary Filtering
  11. Impact of Noise, Drift, and Delay on IMU-Based Estimation
  12. Theory of Complementary Filtering
  13. Kinematic Modeling of the Flexible Manipulator Using the Filtered Joint Angles
  14. Filter Parameter Optimization Using PSO
  15. AI-Augmented Modeling for Residual Correction Using RBFNN
  16. ...and 9 more sections

Figures (7)

  • Figure 1: Flexible manipulator, modeled as a chain of rigid segments, with the corresponding frames.
  • Figure 2: Schematic view of the link $i$ and $i-1$ and their IMUs.
  • Figure 3: Block diagram of the complimentary filtering
  • Figure 4: Experimental setup of the flexible manipulator.
  • Figure 5: Experimental results with no payload under two operating conditions: $\omega = 0.1$ rad/s (first interval) and $\omega = 0.2$ rad/s (second interval).
  • ...and 2 more figures