Design of a Low-Cost Real-Time Angle Sensor Module for Dynamic Body-Weight Support Systems Using Inertial Measurement Unit and Fusion Algorithms
- 3 June 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 26 (13) , 20304-20316
Abstract
Dynamic Body Weight-Support (DBWS) systems are used for patients with mobility impairments. These systems provide three key functions—dynamic body-weight support, fall prevention, and synchronized propulsion—all of which were designed and implemented in this study. One of the most important parameters in a DBWS system is the angle between the patient and the device, which is measured by an angle-estimation module. The problem addressed in this research is the design and construction of a low-cost angle measurement module that is minimally sensitive to vibration, maintains high precision, and has a fast response time. This module is suitable for rehabilitation applications (for example DBWS system) that require high precision and are subject to significant vibration. In this paper, by using an inertial measurement unit (IMU) sensor and a switching logic between different filters based on data fusion, low sensitivity to vibration (vibrations caused by mechanical oscillations) and high response speed are achieved. Outputs of Madgwick and Kalman filters (MF and KF) are fused to obtain the angle from the IMU. The MF is employed to detect rapid movements, and two KFs with high and low measurement noise values (HMNV-KF and LMNV-KF) are used to measure the angle. When rapid angular changes occur, the output angle is obtained from the LMNV-KF; otherwise, the HMNV-KF is used. Additionally, a low-pass filter is applied to the LMNV-KF output after fast angular changes have ended but before the HMNV-KF reaching its steady value. The results obtained from the DBWS system show that the proposed algorithm outperforms all other filters in terms of response speed and vibration sensitivity, making it highly suitable for the intended application. The rise time is 171.33 ms, and the maximum error is 0.54°, both of which are lower than those of all other tested filters.Keywords
This publication has 44 references indexed in Scilit:
- Overground, Body Weight Supported Gait Training Robots: A Scoping ReviewIEEE Access, 2025
- Real-time angle estimation in IMU sensors: An adaptive Kalman filter approach with forgetting factorMechatronics, 2024
- IMU-Based Motion Capture Using Madgwick Filter with 3D Visualization for Robot Teleoperations2024
- Optimal vehicle position estimation using adaptive unscented Kalman filter based on sensor fusionMechatronics, 2024
- Mobility and balance rehabilitation in multiple sclerosis: A systematic review and dose-response meta-analysisMultiple Sclerosis and Related Disorders, 2022
- Fall prevention in older people: past, present and futureAge and Ageing, 2022
- A Smart Mobile Robotic Exoskeleton by Man-Machine Interaction Signal Processing2021
- A Proof of Concept Study for the Design, Manufacturing, and Control of a Mobile Overground Gait-Training SystemInternational Journal of Fuzzy Systems, 2021
- A Novel Adaptive Fading Kalman Filter-Based Approach to Time-Varying Brain Spectral/Connectivity Analyses of Event-Related EEG SignalsIEEE Access, 2020
- Design and development of an adjustable angle sensor based on rotary potentiometer for measuring finger flexion2016