Decentralized Control Design

In this paper, we present a decentralized control strategy which is based on discrete-time adaptive control, for control of the ankle joint in paraplegic subjects using functional electrical stimulation .Agonist-antagonist co-activation is used to control the ankle movement. To achieve this purpose, first, the human is modeled as a single segment inverted pendulum which rotates about the ankle joint. Second, the nonlinear relationship between inclination angle and center of pressure is modeled. Finally, two discrete-time adaptive controllers are used to stabilize the upright posture. Each muscle-joint complex is considered as a subsystem, and separated controllers are designed for each one. Each controller operates individually on its associated

The MIMO control design of such a system requires a complicated mathematical model of musculoskeletal dynamics in the control law design, so to reduce the complexity of the controller a decentralized control scheme is used. The decentralized control problem is to design a set of independent controllers in which each subsystem is controlled by an independent controller. The interaction between the subsystems is taken as external disturbances for each isolated subsystem

In response to backward and forward sway, the ankle flexor and extensor is activated, respectively, so it is required to maintain stability using FES of both flexor and extensor muscles. Here, the control inputs are COP and COP' (velocity of COP), and the agonist (antagonist) muscle joint is a nonlinear time-variant model unlike the control inputs and muscle-joint models that have been proposed in the previous studies.

MATERIALS AND METHODOLOGY

Virtual patient model

A musculoskeletal model is used here as a virtual patient model. The virtual patient model consists of skeletal and muscular parts, so all dynamics are considered.

Inverted pendulum models can be used to explore how CNS controls balance [3]. For skeletal section the human body is approximated and simplified as a single segment, single joint inverted pendulum that rotates about the ankle joint [20]. One reason for this choice of the model is that experimental observations suggest that, for small postural deviations, there is very little (ankle strategy), if any, knee and hip angular motion (i.e., the knee and hip are locked, and they are in full extension.) Single joint inverted pndulum.

The dynamical equation for the inverted pendulum is given by