Identification of Postural Controllers in Human Standing Balance

Standing balance is a simple task for healthy humans but it is not known which control laws are used by the central nervous system (CNS). Identification methods have been used by numerous studies and many mathematical models of the CNS have been extracted, however, limitations exist in these commonly used identification methods. In this chapter, we propose that the trajectory optimization with direct collocation method can identify parametric CNS models from long duration motion data without assuming the identifying system to be linear. We first examined this identification method using synthetic motion data which showed that it can extract correct control parameters. Then, six types of controllers, from simple linear to complex nonlinear, were identified from 100 seconds experimental data. Results from the identifications showed that time-delay and nonlinear property are both needed in order to explain the standing balance motions under randomly external perturbations.