Open-Sim Moco

Open-Sim Moco solves optimal control problems for musculoskeletal models defined in OpenSim to compute optimal motions, muscle activations, and device parameters for biomechanical analysis.

Key Features:

• Direct Collocation Methodology: Employs the direct collocation method to transcribe and solve optimal control problems for musculoskeletal models.
• Handling of Kinematic Constraints: Incorporates kinematic constraints directly into the optimization to enforce model-specific joint and contact relationships.
• Supported Problem Classes: Supports motion tracking, motion prediction, parameter optimization, model fitting, electromyography-driven simulation, and device design.
• Integration with OpenSim: Operates on musculoskeletal models defined in the OpenSim modeling environment.

Scientific Applications:

• Motion Tracking and Prediction: Tracks observed motions or predicts new movements by optimizing muscle activity and joint loading, including replication of walking.
• Movement Prediction (squat-to-stand): Predicts squat-to-stand movements by computing coordinated muscle activations and kinematics.
• Parameter Optimization and Model Fitting: Optimizes model parameters and fits models to data, for example tuning stiffness and parameters of passive assistive devices such as knee braces.
• Electromyography-Driven Simulation: Performs simulations driven by electromyography data to analyze muscle activation patterns during activities.
• Assistive Device Design and Evaluation: Optimizes and evaluates passive assistive devices with objectives related to muscle excitations and knee joint loading.

Methodology:

Solves optimal control problems by transcribing them with the direct collocation method applied to OpenSim musculoskeletal models, using objective functions such as minimizing muscle excitations and knee joint loading as demonstrated in walking replication, squat-to-stand prediction, and passive knee brace optimization.