Abstract
Human motor control is highly efficient in generating accurate and appropriate motor behavior for a multitude of tasks. This paper examines how kinematic and dynamic properties of the musculoskeletal system are controlled to achieve such efficiency. Even though recent studies have shown that the human motor control relies on multiple models, how the central nervous system (CNS) controls this combination is not fully addressed. In this study, we utilize an Inverse Optimal Control (IOC) framework in order to find the combination of those internal models and how this combination changes for different reaching tasks. We conducted an experiment where participants executed a comprehensive set of free-space reaching motions. The results show that there is a trade-off between kinematics and dynamics based controllers depending on the reaching task. In addition, this trade-off depends on the initial and final arm configurations, which in turn affect the musculoskeletal load to be controlled. Given this insight, we further provide a discomfort metric to demonstrate its influence on the contribution of different inverse internal models. This formulation together with our analysis not only support the multiple internal models (MIMs) hypothesis but also suggest a hierarchical framework for the control of human reaching motions by the CNS.
Highlights
Mechanisms along with learning and adaptation processes for motor control has been studied extensively[11,12,13,14,15,16]
Inverse Optimal Control (IOC) formulation describes human motion better than the previous single models, yet it forms a more complex computational problem, which emphasizes the importance of a better understanding of why and how those models are efficiently utilized by the central nervous system (CNS)
Considering that internal models mimic the transformations between system states, motor commands, and sensory signals, an optimal control problem (OCP) can be regarded as an inverse internal model by providing the necessary control signals to carry out the reaching task[2,17,36]
Summary
Mechanisms along with learning and adaptation processes for motor control has been studied extensively[11,12,13,14,15,16]. Recent studies focus on finding a combination of such models These methods solve an Inverse Optimal Control (IOC) problem where the contribution of different optimal control models are computed[33,34,35]. For the arm reaching task we focused, a single forward model, i.e. the arm dynamics, is paired with multiple inverse internal models, i.e. the composite of OCP where each OCP is associated with a specific cost function. This paper presents both a comprehensive optimality analysis on human arm reaching motions in 3D space and a trade-off between dynamics and kinematics based controllers depending on the reaching motion type. Multiple inverse internal models, represented as a combination of different controllers, are discovered to be collectively controlling the 3D free-space arm reaching movements
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
