| Summary: | With the elderly and disabled population increasing worldwide, the functionalities of smart wheelchairs as mobility assistive equipment are becoming more enriched and extended. Although there is a well-established body of literature on fatigue detection methods and systems, fatigue detection for wheelchair users has still not been widely explored. This paper proposes a neuro-fuzzy fatigue tracking and classification system and applies this method to classify fatigue degree for manual wheelchair users. In the proposed system, physiological and kinetic data are collected, including surface electromyography, electrocardiography, and acceleration signals. The necessary features are then extracted from the signals and integrated with a self-rating method to train the neuro-fuzzy classifier. Four degrees of fatigue status can be distinguished to provide further fatigue and alertness prediction in the event of musculoskeletal disorders caused by underlying fatigue.
|
|---|
