In this paper we present a fuzzy reinforcement learning control approach that requires no previous knowledge about vehicle model characteristics. By means of simulations we show that the scheme can perform well under a variety of maneuver and road conditions and adapt its behavior accordingly without requiring any overly complicated operations. Rule based fuzzy systems have been extensively applied with success in this area due to their similarity to human perception and reasoning. The fuzzy rules are optimized by reinforcement learning algorithms. The output of this procedure is an optimal policy which controls the vehicle dynamic stability.

Keywords