In this study, we propose an estimation method of physical parameters of the plants and their uncertainty of the controlled object using the system identification and Unscented Transformation (UT). By accurately estimating the uncertainty, the deterioration of the precision of the state estimate is expected to be reduced.