Abstract
A Study of comparison between kalman filter and human controller's prediction characteristics.
Kalman Filter has been often used as a mathematical model of human controller's estimation function, but it is not clear in reality whether or not he estimates the movement of a target according to Kalman Filter function.
In this paper, we have tried to set up an experimental situation of two-variables prediction in which only the estimation characteristics of human controller coulo be isolated, and compared the experimental findings with the theoretical model fd Kalman Filter. Our results show that (1) Kalman Filter model fits well to the estimation characteristics of human controller and that (2) he uses first and second differential values of target as the important information for his estimation.
