“…as input variables, and uses neural networks [ 11 , 12 , 13 ], fuzzy rules [ 3 , 14 , 15 ], genetic algorithm [ 16 ], deep learning [ 17 , 18 , 19 ], etc., to establish nonlinear mapping models of the parameters to be estimated. Alternatively, the least squares method [ 20 , 21 , 22 ], Bayesian estimation [ 23 ], a Kalman filter [ 21 , 24 , 25 , 26 ], or other estimation criteria can be used to identify parameters, estimating the body mass, body side angle, road slope, slip rate, adhesion coefficient, driving style, and road roughness.…”