Empowering human-like walking with a bio-inspired gait controller for an under-actuated torque- driven human model

Abstract

Human gait simulation plays a crucial role in providing insights into various aspects of locomotion, such as diagnosing injuries and impairments, assessing abnormal gait patterns, and developing assistive and rehabilitation technologies. To achieve more realistic results in gait simulation, it is necessary to utilize a comprehensive model that closely replicates the kinematics and kinetics of the human gait pattern. OpenSim software provides anthropomorphic and anatomically accurate human skeletal structures that enable users to create personalized models for individuals to accurately replicate real human behavior. However, torque-driven models face challenges in balancing unactuated degrees of freedom during forward dynamic simulations. Adopting a bio-inspired strategy that ensures an individuals balance with a minimized energy expenditure, this paper proposes a gait controller for a torque-deriven OpenSim model to achieve a stable walking. The proposed controller takes a model-based approach to calculate a "Balance Equivalent Control Torque" and uses the concept of the hip-ankle strategy to distribute this balance torque to the lower-limb joints. To optimize the controller gains and the "Balance Distribution Coefficients", an interface is stablished between MATLAB and OpenSim that is capable of conducting controllable forward dynamic simulations. The simulation results demonstrate that the torque-driven model can walk naturally with joint torques suitably matching experimental data. The robustness of the bio-inspired gait controller is also assessed by applying a range of external forces on the upper body to disturb the model. The robustness analysis demonstrates the quick and effective balance recovery mechanism of the proposed bio-inspired controller.