The rising demand of autonomous and intelligent transportation systems requires the development of advanced control and estimation techniques, aiming to ensure safety and efficient operations. Due to the nonlinear nature of vehicle dynamics and its characteristic phenomena, classical estimation and control methods may not achieve adequate results, which encourages the research of novel algorithms. By some contributions, the first part of this work deals with estimation algorithms, both for identification of time invariant parameters and for estimation of states and time varying parameters. Special emphasis is given to Moving-Horizon State Estimation (MHSE), which is presented to be robust and accurate, due to the constrained optimization problem on which it is based. This algorithm is evaluated in vehicle longitudinal dynamics, for slip and tire-road friction estimation. Despite its efficiency, the high computational cost makes it necessary to search for suboptimal alternatives, and the employ of a Neural Networks that maps the optimization results is a promising solution, which is treated as Neural Networks Moving-Horizon Estimation (NNMHE). The NNMHE is evaluated on a state-of-charge (SOC) estimation of batteries for electric vehicles, demonstrating, through experimental data, that the NNMHE emulates accurately the optimization problem, and the literature indicates its effectively application on embedded hardware. Finally, a contribution about Nonlinear Model-based Predictive Control (NMPC) is presented. It is proposed and evaluated its use compounding a novel hierarchical control framework for electric vehicles with independent in-wheel motors, through which it is possible to adequately control the vehicle on velocity and path tracking tasks, with reduced computational effort. The control is evaluated using experimental obtained tire data, which approaches the simulation to real situations.