Human model in the loop design optimization for RoboWalk wearable device

In this paper, a design optimization approach for assistive devices based on the simulation of the robot and human model is proposed. The proposed human model in the loop (HMIL) approach exploits an augmented human-robot model for interaction analysis, and to evaluate and modify the assistive device effectiveness. In this regard, a lower-limb assistive device (RoboWalk) for elderly and a human skeletal system are modeled and augmented to use the human-robot interaction forces in the proposed HMIL strategy. Due to the particular design of RoboWalk, a complete set of constraints are specified to ensure the solution feasibility and compatibility of the device in different poses. This constrained optimization problem is solved by weighted sum particle swarm optimization (WSPSO) and multi-objective PSO (MOPSO) algorithm. As a result of this optimization, several appropriate sets of design parameters are found to minimize the load on human joints, actuator torque and mass. Among the obtained optimal points, the point minimizing an introduced index is selected as the final design.