Vehicle State Estimation Based on Recurrent Neural Network

For vehicle state estimation, a new method based on recurrent neural network is proposed to improve the accuracy of the estimation and to simplify the algorithm. The method not only eliminates the necessity of constructing complex vehicle models, but is also suitable for non-linear systems and has good convergence. Through simulations under different operating conditions, the results show that this new method can accurately estimate many parameters that are difficult to estimate, such as the vehicle's sideslip angle, road adhesion coefficient, etc.