A direct-drive linear actuated assistive robot for the shoulder

Robot-assisted approaches for upper limb rehabilitation after stroke have many advantages over conventional therapy. However, the majority of current robots have not been shown to improve the outcome of therapy significantly. One of the main sources of their ineffectiveness is that these robots utilise heavy and bulky industrial motors and often require a complicated mechanism to provide the required motion for the patient’s shoulder. Therefore, the majority of these robots are stationary and are unable to provide a meaningful task-specific rehabilitation. To increase the effectiveness of therapy, they need to become like assistive robots: wearable, portable, and adaptable. Direct-drive linear synchronous motors can potentially solve many challenges faced in the use of these robots. However, off-the-shelf motors are unable to meet the requirements for rehabilitation robots. This thesis explores the application of permanent magnet linear synchronous motors (PMLSM) in designing a "one-size-fit-all" shoulder assistive robot that can provide gravitational support as well as repetitive task-specific training. Using a derived mathematical model of the robot-human interaction, the optimum location of the robot on the human body was found. Combining a thermal and a semi-analytical electromagnetic model of the motor, the motor’s total mass, including its cooling system, was optimised. The motor which fits best the requirements for the robot was chosen for manufacturing. Using the result of the optimisation, a motor prototype was constructed and incorporated into the assistive robot. The fabricated robot had a total mass of 1.3 kg with nominal continuous force and power of 91N and 80W respectively. The motor used an array of copper fins and a high-pressure DC fan for its cooling system. The motor constant was measured to be 9.21N/ √ W suggesting that the motor could achieve a high continuous force density of ∼ 128N/kg and produce the required force for the assistive robot. The fabricated robot, which was subjected to a standard safety assessment, was tested on healthy individuals to investigate the effectiveness of the robot’s gravity compensatory assistance and resistance on the average activities of the shoulder muscles. For this study, two control strategies were designed for the robot: a PID position controller for total assistance and a current controller to provide fixed partial assistance/resistance. Both assistive strategies and resistive strategies significantly changed the activity of the majority of shoulder muscles across different motions (P < 0.05). Moreover, significant task-specific correlations were observed between the muscle activities and the robot’s level of assistance/resistance for most of the muscles (P < 0.05). The maximum effect of the robot was on the deltoid muscle with ∼ 70% and 66% reductions in average muscle activity during vertical trajectory and isometric abduction tasks, respectively, when the robot was providing total assistance. Based on the human study results, the robot can potentially be considered a valuable rehabilitation tool to reduce muscular activities and to target muscle strength. Moreover, the design methodology explained in the thesis is applicable to the design of actuators for other types of assistive robots and opens the door to increased capability and effectiveness of assistive and rehabilitation robots.

Name thesis:degree_name

PhD

Level thesis:degree_level

Doctoral

Discipline thesis:degree_discipline

Bioengineering

Grantor dc:publisher

ResearchSpace@Auckland

Year dc:date.issued

2021

Author dc:creator

• Haji Hosseinnejad, Soroosh

Advisors dc:contributor.advisor

• Ruddy, Bryan
• Besier, Thor
• Taberner, Andrew

Statement dc:rights

• Items in ResearchSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Licence dc:rights.uri

• https://researchspace.auckland.ac.nz/docs/uoa-docs/rights.htm

Handle dc:identifier.uri

https://hdl.handle.net/2292/61737

OAI identifier oai:identifier

oai:researchspace.auckland.ac.nz:2292/61737