Detection of abrupt changes in autonomous system fault analysis using spatial adaptive estimation of nonparametric regression

The paper deals with the detection of abrupt changes in autonomous systems. We consider this problem in the presence of Gaussian noise and solve it in two steps. At first, spatial adaptive estimation of nonparametric regression is used to estimate the observable data. Then Filtered Derivative Algorithm is used to detect abrupt changes in the obtained data using an adaptive threshold. The estimation of this adaptive threshold is presented. This approach is then applied to demonstrate the slowdown detection of a small autonomous vehicle.