Tampere University of Technology

TUTCRIS Research Portal

Inertial Sensor-Based State Estimation of Long-Reach Flexible-Link Manipulators

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Details

Original languageEnglish
Title of host publication2017 IEEE 8th International Conference on Cybernetics and Intelligent Systems (CIS) and IEEE Conference on Robotics, Automation and Mechatronics (RAM)
PublisherIEEE
Number of pages6
ISBN (Electronic)978-1-5386-3135-5
DOIs
Publication statusPublished - 2018
Publication typeA4 Article in a conference publication
EventIEEE International Conference on Cybernetics and Intelligent Systems (CIS) and IEEE Conference on Robotics, Automation and Mechatronics (RAM) -
Duration: 1 Jan 2000 → …

Conference

ConferenceIEEE International Conference on Cybernetics and Intelligent Systems (CIS) and IEEE Conference on Robotics, Automation and Mechatronics (RAM)
Abbreviated titleCIS-RAM
Period1/01/00 → …

Abstract

In this paper, we study the performance of a finite-element-based observer in estimating the flexural degrees of freedom (DOF) of a single-link flexible long-reach manipulator. The observability of the system confirms that all the flexural states of the system can be estimated using a single angular velocity measurement. The inputs to the observer are obtained from retrofittable and low-cost inertial sensors suitable for mobile machines. Results of the observer's performance with different measured states and varying loads are provided. Validation of the end-point position is carried out using an OptiTrack camera system. The performance of the estimated variables as feedback signals in high-performance control is demonstrated using a nonlinear model-based controller based on the virtual decomposition control framework. The test setup consists of a hydraulically driven 4.5-meter-long beam having a maximum tip mass of 70 kg, resulting in a static deflection of -56.7 cm. The initial experiments on the 1-DOF system indicate that the proposed method is effective.

Publication forum classification