Abstract
In the case of driver assist systems that can be accepted by the driver, it need to make clear the control performance of the driver by the driver's model. This research did modeling of the decelerating action by the Fuzzy Neural Network (FNN). The input data divided the learning data and the non-learning data. It was learned only the learning data. The simulation of the decelerating action did about the non-learning data. As a result, it is possible for FNN to do the decelerating action that approximate to the driver's action.