Neural Network Observer for Lateral Vehicle Model with Varying Sampled and Delayed Output

This research offers a neural network adaptive observer (NNAO) architecture for nonlinear lateral vehicle dynamics with variable sampled delayed output. A radial basis function (RBF) neural network is used to approximate the system's unknown part, and a new weight updating mechanism is provided. A closed-loop output predictor is used to offer inter-sample output estimate while dealing with variable samples, and a closed-loop integral compensation is used to deal with variable delay. The convergence of the proposed observer is proved using Lyapunov function and small gain arguments. Simulation tests confirm the NNAO's estimate algorithm's accuracy in estimating yaw rate, longitudinal speed, and particularly the excellent performance of the estimation of lateral speed.