The constantly growing impact of neurological diseases on world population has led in the last years to a constantly increasing interest and research in the field of rehabilitation technologies. In this scenario, hybrid systems combining powered exoskeletons and Functional Electrical Stimulation (FES) are the most studied due to the advantages given by the coordinated assistance to the user’s muscles during the execution of a functional movement. The combination of these two contributions helps exploiting their respective advantages and limit their drawbacks. However, most hybrid systems lack a fully cooperative control between motor-actuated robot contribution, FES contribution and residual voluntary movements of the patient, which is needed to maximize the rehabilitation outcomes and favor neural plasticity. To address these challenges, in this work an Electromyography(EMG)-triggered control framework for hybrid rehabilitation systems was developed, capable of integrating the three contributions: motor (MOT), FES and voluntary (VOL). The motor torque contribution is regulated by an impedance control, which accounts for weight compensation in feed-forward and for trajectory error correction in feed-back; the charge delivered with FES is regulated by means of an Iterative Learning Control, based on the tracking error of the previous repetition executed. To promote user’s involvement, thus avoiding a slacking phenomenon, the FES and motor contribution is triggered by the user’s volitional muscle contractions and are consequently regulated also taking into account the user’s voluntary contribution to the exercise. This cooperative control was tested using a motor-actuated robotic test-bench for the knee flexion-extension exercise on 10 healthy subjects. Results showed that, when comparing the fully hybrid solution (FES+MOT+VOL) with the execution of passive motor-driven exercises, a decrease of 94% in torque power demand is found, without any significant decline in tracking performance. Moreover we could also appreciate a decrease in charge (of 68%) and torque (of 60%) demand in the FES+MOT+VOL modality with respect to the FES+MOT modality, which is coherent with the introduction of a third contribution (the voluntary effort) to the exercise. Future developments will involve monitoring EMG activity also during the exercise to provide real-time feed-back on users’ active participation and to improve their engagement and transferring this control strategy on a multi-joint exoskeleton.
Il costante aumento dell’impatto di malattie neurologiche sulla popolazione mondiale ha condotto negli ultimi anni a un interesse crescente e alla ricerca nell’ambito delle tecnologie riabilitative. In questo contesto, sistemi ibridi che combinano esoscheletri attivi e Stimolazione Elettrica Funzionale (FES) sono i maggiormente studiati grazie ai vantaggi dati dall’assistenza coordinata che forniscono ai muscoli dell’utente durante l’esecuzione di movimenti funzionali. La combinazione di questi due contributi aiuta a sfruttare i rispettivi vantaggi ed allo stesso tempo limitare i rispettivi svantaggi. La maggior parte dei sistemi ibridi presenti allo stato dell’arte non possiedono un controllo completamente cooperativo tra il contributo dato dal motore che attua il robot, il contributo della FES e il contributo volontario residuo dell’utente, che sarebbe invece necessario per massimizzare i risultati della riabilitazione aumentando la plasticità neurale. Per andare a colmare questa lacuna, nel presente lavoro è stato sviluppato un sistema di controllo per sistemi riabilitativi ibridi triggerato dall’attivazione elettromiografica (EMG) volontaria dell’utente che andasse ad integrare i tre contributi: motore (MOT), FES e volontario (VOL). La coppia fornita dal motore è regolata tramite un controllo in impedenza, che compensa il peso in feed-forward e gli errori in traiettoria in feed-back. La carica fornita dalla FES è regolata da un controllo iterativo (Iterative Learning Control, ILC) basato sull’errore in traiettoria della ripetizione precedente. Nell’ottica di promuovere la partecipazione dell’utente (e di evitare il fenomeno dello slacking), il contributo del motore e della FES sono triggerati dalle contrazioni volontarie del muscolo dell’utente, e consecutivamente regolati prendendo in considerazione il contributo volontario dell’utente durante l’esercizio. Questo controllo cooperativo è stato testato utilizzando un banco prova robotico attuato da un motore. Il movimento indagato è stato la flesso-estensione di ginocchio di 10 soggetti sani. I risultati hanno mostrato che, paragonando la modalità completamente ibrida (FES+MOT+VOL) con la modalità in cui l’intero esercizio è eseguito solamente dal motore (MOT), si ha un decremento nella potenza richiesta al motore pari al 94%, senza che questo incida consistentemente sul tracciamento della traiettoria desiderata. Inoltre, nella modalità ibrida FES+MOT+VOL si è potuto apprezzare un decremento nella quantità di carica (pari al 68%) e nella quantità di coppia (pari al 60%) rispetto alla modalità ibrida FES+MOT, risultato coerente con l’introduzione del terzo contributo (volontario) nell’esercizio riabilitativo. Gli sviluppi futuri del presente lavoro comprendono: il monitoraggio dell’attivazione EMG in tempo reale durante l’esercizio per fornire un feed-back all’utente sul suo livello di partecipazione al fine di migliorarne il coinvolgimento; il trasferimento del sistema di controllo cooperativo presentato su un esoscheletro con più di un grado di libertà.
An EMG-triggered cooperative controller for a single-joint hybrid FES-robotic system
Zimei, Eva
2022/2023
Abstract
The constantly growing impact of neurological diseases on world population has led in the last years to a constantly increasing interest and research in the field of rehabilitation technologies. In this scenario, hybrid systems combining powered exoskeletons and Functional Electrical Stimulation (FES) are the most studied due to the advantages given by the coordinated assistance to the user’s muscles during the execution of a functional movement. The combination of these two contributions helps exploiting their respective advantages and limit their drawbacks. However, most hybrid systems lack a fully cooperative control between motor-actuated robot contribution, FES contribution and residual voluntary movements of the patient, which is needed to maximize the rehabilitation outcomes and favor neural plasticity. To address these challenges, in this work an Electromyography(EMG)-triggered control framework for hybrid rehabilitation systems was developed, capable of integrating the three contributions: motor (MOT), FES and voluntary (VOL). The motor torque contribution is regulated by an impedance control, which accounts for weight compensation in feed-forward and for trajectory error correction in feed-back; the charge delivered with FES is regulated by means of an Iterative Learning Control, based on the tracking error of the previous repetition executed. To promote user’s involvement, thus avoiding a slacking phenomenon, the FES and motor contribution is triggered by the user’s volitional muscle contractions and are consequently regulated also taking into account the user’s voluntary contribution to the exercise. This cooperative control was tested using a motor-actuated robotic test-bench for the knee flexion-extension exercise on 10 healthy subjects. Results showed that, when comparing the fully hybrid solution (FES+MOT+VOL) with the execution of passive motor-driven exercises, a decrease of 94% in torque power demand is found, without any significant decline in tracking performance. Moreover we could also appreciate a decrease in charge (of 68%) and torque (of 60%) demand in the FES+MOT+VOL modality with respect to the FES+MOT modality, which is coherent with the introduction of a third contribution (the voluntary effort) to the exercise. Future developments will involve monitoring EMG activity also during the exercise to provide real-time feed-back on users’ active participation and to improve their engagement and transferring this control strategy on a multi-joint exoskeleton.File | Dimensione | Formato | |
---|---|---|---|
Executive_Summary___Eva_Zimei.pdf
accessibile in internet per tutti
Dimensione
1.1 MB
Formato
Adobe PDF
|
1.1 MB | Adobe PDF | Visualizza/Apri |
Tesi_formato_articolo___Eva_Zimei.pdf
accessibile in internet per tutti
Dimensione
1.67 MB
Formato
Adobe PDF
|
1.67 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/208644