Unilateral Lower Extremity Exoskeleton Utilizing Sensor Fusion Algorithms

Dizor, R.; Raj, A.; Stewart, T.; Gonzalez, B.; Smith, G.; Carter, Z.; Domingues, B.; Newton, J.

doi:10.1109/rapid54473.2023.10264748

Cited by 1 publication

(3 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The deployment of two sEMG/IMU sensor arrays on the thigh, one anterior over the quadriceps and vastus muscles and another posterior over the hamstrings, coupled with an additional array around the calf and mid-shin and an IMU array on the footplate, exemplifies a comprehensive approach to capturing a full spectrum of lower limb movements. Each sensor array is a repurposed Thalmic Labs armband originally meant for arm and hand gestures to control other devices wirelessly [4,9]. This array configuration ensures the detection and interpretation of complex muscle activities and movements necessary for executing a range of ADLs [11].…”

Section: Sensor Designmentioning

confidence: 99%

“…A sensor fusion algorithm combines the data real-time to classify the intended movement intention and joint position of the user and the exoskeleton [4]. Leveraging the sensor data, the MCU based sliding mode control offers a sophisticated mechanism for dynamically adjusting air pressure in the PAMs, optimizing the exoskeleton's movements for smoothness and natural alignment with the user's intentions.…”

Section: Sensor Designmentioning

confidence: 99%

“…Our focus lies in employing deep reinforcement learning algorithms for real-time, adaptive control. Central to our approach is the use of sensor fusion, combining sEMG and IMU data to create a detailed real-time profile of muscle activity and limb movement as well as positioning [4]. The LSTM network is pivotal in interpreting these real-time sequential data streams, enabling the prediction of user intentions by analyzing patterns over time [20].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Deep reinforcement learning to assess lower extremity movement intention and assist a rehabilitation exoskeleton

Dizor,

Raj,

Gonzalez

et al. 2024

Disruptive Technologies in Information Sciences VIII

View full text Add to dashboard Cite

This paper introduces a pioneering approach for controlling a unilateral lower extremity exoskeleton designed for rehabilitation and enhancing the quality of life for individuals with neuromuscular weakness of the lower limbs. At the core of our methodology is the integration of Long Short-Term Memory (LSTM) networks with Proximal Policy Optimization (PPO) models, utilizing a deep reinforcement learning framework to interpret and predict user movement intentions in real time. By harnessing sensor fusion that combines surface electromyography (sEMG) and inertial measurement units (IMU) from sensor arrays placed around the quadriceps and gastrocnemius muscles, our system employs an adaptive nonlinear sliding mode control with pneumatic artificial muscles (PAMs), thereby directing the exoskeleton's movement and positioning. The LSTM network processes temporal sequences of sensor data to capture the dynamics of human motion, while the PPO model optimizes the control policy to ensure smooth and responsive movements aligned with the user intentions. Focusing initially on basic maneuvers integral to Activities of Daily Living (ADL), our system demonstrates promising preliminary results in mimicking natural limb movements, laying the groundwork for future clinical applications. This paper specifically delves into the utilization of the LSTM-PPO framework for controlling an avatar prior to testing the exoskeleton, representing a significant step towards realizing a responsive and intuitive exoskeleton control system.

show abstract

Section: Sensor Designmentioning

confidence: 99%

Section: Sensor Designmentioning

confidence: 99%