An exoskeleton arm optimal configuration determination using inverse kinematics and genetic algorithm

Area of Science:

• Robotics
• Biomechanics
• Rehabilitation Engineering

Background:

• Human arm biomechanics are complex, requiring detailed kinematic models for effective exoskeleton design.
• Existing rehabilitation exoskeletons may lack optimal configurations tailored to individual patient needs and activities.

Purpose of the Study:

• To develop a kinematic model for an arm exoskeleton for human rehabilitation.
• To determine optimal exoskeleton arm configurations based on patient capabilities and user activity.

Main Methods:

• Utilized Denavit-Hartenberg notation for upper arm modeling.
• Developed the exoskeleton-human arm model in MathWorks.
• Employed multicriteria optimization and artificial intelligence, specifically genetic algorithms, for trajectory planning.

Main Results:

• Optimal exoskeleton arm configurations were identified using a genetic algorithm.
• Movement trajectories and joint angle changes were visualized for two motion variants.
• Pareto-optimum solutions for movement trajectories were presented, with a utopia point enabling single solution selection.

Conclusions:

• The proposed kinematic modeling and genetic algorithm approach is efficient for analyzing exoskeleton kinematics and dynamics.
• Genetic algorithms offer a viable solution for commanding actuators in multi-criteria optimization problems.
• Future work should involve experimental validation of the proposed method's effectiveness.