The current trend to completely replace the human driver in vehicles is actually a challenge for control engineering. A fully autonomous vehicle, which qualifies for the fifth level of driving automation according to SAE, includes a symbiotic system of control functions, so that the car contains an aggregate of interdependent controllers that process signals and data about the state of the vehicle and its interaction with the environment in real time. Many of these controllers are automatic systems with the basic function of a regulator, some are systems based on decision rules, and others are endowed with a higher level of intelligence, being able to learn and deal with situations of uncertainty. An artificial copy of the human driver must be a mix of response models and decision strategies that provide timely, optimal and safe actions. In this paper, classical and advanced control models are discussed and of course those based on AI in which the learning function and logical inference mechanisms are specifically described. Control strategies are those that define the autonomous behavior of the vehicle, a higher-level macrocontrol, which gives nuance to decisions leading ultimately to responsible actions on the vehicle. The paper analyzes the control models and strategies for the main subsystems of a vehicle within simulated scenarios. The content of the paper is useful for the better development of autonomous vehicles by challenging the design of safe and less expensive control architectures.