Information rate of human manual control of unstable systems

Human operators are active components of feedback control systems. They are faced with challenges especially when the system is unstable or when time delay affects the feedback loop. In this study, the human operator is assumed to act as a dynamical system, and its complexity and performance is analyzed through the entropy rate of its generated output command. We investigated the characteristics of the human operator when controlling an unstable system under different time delays and higher degrees of task instability. The entropy rate for dynamical systems, also called the Kolmogorov-Sinai entropy, was estimated based on the numerical method proposed by Grassberger and Procaccia.