Optimal control modeling of human movement

Abstract

Optimal control theory is a common paradigm employed in many fields of science and engineering. This chapter provide an overview of optimal control theory applied specifically to studying the biomechanics and control of human movement. In this approach, the human neuromusculoskeletal system is modeled as a system of ordinary differential equations subject to controls that influence the behavior of the system. Techniques from control theory are used to find the optimal controls that cause the model to behave in a manner that minimizes or maximizes a user-defined performance criterion. The performance criterion, along with any task requirements, mathematically define the goal of the movement to be simulated. This framework has proven to be both powerful and flexible, leading to fundamental insights on topics ranging from the biomechanical function of individual muscles in locomotion to the manner in which the nervous system controls limb movements in reaching tasks. Many of the basic concepts that are introduced in the first half of this chapter are then demonstrated via a detailed example of the optimal control of human walking. Selected contemporary topics that hold promise for the future are discussed, as are challenges with the use of optimal control theory. The chapter concludes with perspectives on future developments in the application of optimal control theory to human movement.