Adaptive Trajectory Generation For Vision-Based Robot Using Negotiation Principle For Rehabilitation Applications

• Main Author: Rasid, Farah Amirah
• Format: Thesis
• Language: English; English
• Published: 2019
• Subjects: TJ Mechanical engineering and machinery
• Online Access: http://eprints.utem.edu.my/id/eprint/24716/; https://plh.utem.edu.my/cgi-bin/koha/opac-detail.pl?biblionumber=116878

Adaptive behaviour in a robotic system is highly desired in an application that requires a robot to negotiate and adapt its role to the overall goal. For example, in an autonomous hand rehabilitation application, the robot must concern on the safety and comfort of a patient when guiding the rehab exercise and the robot must also be able to motivate patient to gradually improve his motion to achieve recovery. Therefore, this research focuses in solving the problem by adjusting the robot's trajectory automatically to suit changing patient’s requirement while considering the planned trajectory. To solve the problem, it is hypothesized that persuasion tactic based on negotiation principle approach can solve the conflict in motion coordination and lead the patient towards the initial planned trajectory. The objectives for this research are to design and develop an adaptive trajectory algorithm with a persuasion tactic based on negotiation principles and to validate the proposed algorithm in terms of automated negotiation point of view. A persuasion tactic is proposed to represent the action of persuasion. Three types of experiments were conducted to validate the proposed method. The first experiment is to study the effect of using negotiation principle without persuasion tactic on the trajectory of a robot. The second experiment is to investigate the effect of using negotiation principle with persuasion tactic on the trajectory of a robot (single input). The third experiment is to examine the effect of using negotiation principle with persuasion tactic on the trajectory of a robot (multi-input). The result of the experiment is analyzed based on a negotiation perspective in terms of negotiated trajectory, the success rate of negotiation, utility, and equality. The result shows that the success rate of negotiation with persuasion tactic is higher than without persuasion tactic which is 91.1%. The result also shows that the utility of the robot is higher than human which are 0.63 (single input) and 0.52 (multi-input). In conclusion, the presented persuasion tactic based on negotiation principle to coordinate motion between robot and human serves as a method that fills the gap in the robotic system which does not consider its initial planned trajectory and totally follows the human requirement.