Modeling of Functional Electrical Stimulation (FES): Powered Knee Orthosis (PKO) assisted gait exercise in post-stroke rehabilitation / Adi Izhar Che Ani
Hemiplegics can recover or at least regain some function. However, if the patient is not appropriately treated, the consequences of the chronic stage will be long-lasting. Rehabilitation exercise has been deemed one of the most promising methods for hemiplegic regain function. Researchers have recen...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2023
|
| Online Access: | https://ir.uitm.edu.my/id/eprint/88633/1/88633.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Hemiplegics can recover or at least regain some function. However, if the patient is not appropriately treated, the consequences of the chronic stage will be long-lasting. Rehabilitation exercise has been deemed one of the most promising methods for hemiplegic regain function. Researchers have recently utilized various exercise techniques in conjunction with Functional Electrical Stimulation (FES). Rehabilitation exercise using a Powered Knee Orthosis (PKO) has been extensively studied, particularly concerning the practical adaption of the users and control technique for FES. In the various mechanism of actuating the PKO, such as rotary series elastics actuators, dc motor actuators, and brushless dc motor actuators, researchers continue to make control strategies to get better actuating of PKO. However, there are still shortcomings in modeling the rehabilitation system involving FES and PKO. Having FES and PKO modeling for the rehabilitation system is critical for controlling PKO and FES. The current technique uses statistical models, but the dynamic properties of human physiology limit the current approach. This research aimed to propose modeling FES – PKO-assisted gait exercise in post-stroke rehabilitation through sub-modeling parts, which are the human gait model and the FES and PKO model. In the human gait model, three Machine Learning algorithms were used: Gaussian Process Regression, Support Vector Machine, and Decision Tree. The improvement of the model was implemented by incorporating different sliding windows on the model's input parameters. The best gait model was a Decision Tree with sliding window data (t-3), which had a root mean square error of 3.3018 and an R-squared value of 0.97. The human gait model was then integrated with the quadriceps muscle model developed by the previous researcher. The quadriceps muscle model is one of the selected muscle groups. In this physiologically based muscle model, the major properties of the human muscle are described in three components: muscle activation, muscle contraction, and, body segmental dynamic. The next sub-model consists of the PKO model. The dynamic model of PKO is developed by developing mathematical equations that connect the torques exerted by the actuator of PKO. The Lagrange formula is employed here to build the PKO’s inverse dynamic model equations. The complete modeling of the system was validated with our developed prototype of PKO via experimental work. The developed FES-PKO assisted gait exercise using the approach that combines human gait, muscle, and PKO model shows that FES and PKO generated joint moment and joint angle can simulate movement during gait. These developed models contribute to FES as the primary source in generating joint movement while PKO assists as needed. |
|---|