Extremum Seeking-based Adaptive Feedback Control of Functional Electrical Stimulation: Elbow Joint control in Silico
Haukipää, Markus (2025)
2025
Bioteknologian ja biolääketieteen tekniikan maisteriohjelma - Master's Programme in Biotechnology and Biomedical Engineering
Lääketieteen ja terveysteknologian tiedekunta - Faculty of Medicine and Health Technology
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2025-08-18
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202508188299
https://urn.fi/URN:NBN:fi:tuni-202508188299
Tiivistelmä
Stroke significantly impacts upper limb functionality, with up to 70 % of survivors affected. Functional electrical stimulation is a common rehabilitation method. Electrical pulses stimulate the muscles and induce movement. The combination of stimulation and movement helps regain muscle control. Due to the non-linear and inconsistent nature of muscle stimulation, precise and deterministic control is difficult to achieve —despite it being the key to faster recovery.
To improve the control, an adaptive feedback controller is proposed. The controller consists of two parts: a Proportional-Integral (PI) controller and a stochastic extremum seeking algorithm. Electrical stimulation is controlled utilizing the angular error of the elbow joint. Moreover, the algorithm aims to improve responsiveness through time and non-linear changes.
The goal is to form a comprehensive and repeatable testing platform to challenge the adaptive controller’s capability to broaden the understanding of its applications. Owing to the goal, the controller, as well as the testing platform, is completely executed in software. The controller is evaluated in continuous excitation, covering demanding movement patterns and fluctuating non-linear characteristics. The results are compared to traditionally tuned fixed PI controller.
The adaptive controller demonstrated up to 3° of controlled improvement in root-mean-square-error versus the fixed controller. However, the system introduced complications in consistency. This is due to the adaptive controller’s performance being orchestrated by a stochastic process. For practical implementation, reliability must be ensured to guarantee patient safety and performance. For further improvements, the thesis established possible solutions and a pre-made testing environment. With the problems solved, the adaptive controller demonstrated potential in non-linear control. Looking forward, it may present a possible base for future controllers — controlling the adaptation of the future’s healthcare.
To improve the control, an adaptive feedback controller is proposed. The controller consists of two parts: a Proportional-Integral (PI) controller and a stochastic extremum seeking algorithm. Electrical stimulation is controlled utilizing the angular error of the elbow joint. Moreover, the algorithm aims to improve responsiveness through time and non-linear changes.
The goal is to form a comprehensive and repeatable testing platform to challenge the adaptive controller’s capability to broaden the understanding of its applications. Owing to the goal, the controller, as well as the testing platform, is completely executed in software. The controller is evaluated in continuous excitation, covering demanding movement patterns and fluctuating non-linear characteristics. The results are compared to traditionally tuned fixed PI controller.
The adaptive controller demonstrated up to 3° of controlled improvement in root-mean-square-error versus the fixed controller. However, the system introduced complications in consistency. This is due to the adaptive controller’s performance being orchestrated by a stochastic process. For practical implementation, reliability must be ensured to guarantee patient safety and performance. For further improvements, the thesis established possible solutions and a pre-made testing environment. With the problems solved, the adaptive controller demonstrated potential in non-linear control. Looking forward, it may present a possible base for future controllers — controlling the adaptation of the future’s healthcare.