A 2-DOF model of the five-axle all terrain crane with some simplifications is proposed, and based on the assumption of zero side-slip angle of mass center and the proportional control strategy, the function relating the angles of the rear and the front axles, and the velocity are derived. With some simplifications with respect to the proportional valve and the cylinder, a mathematical model of the steering system in Matlab/Simulink is established. In order to overcome the weakness of the traditional PID controller, a fuzzy adaptive PID controller of the rear axle steering angle is built in Matlab by using its fuzzy logic control toolbox. Through the real-time determination of the PID parameters, a high control precision is achieved. A steady state cornering simulation is carried out under 25km/h, and the comparison of the simulation and theoretical values shows that the fuzzy adaptive PID controller works well, with quick system response and good following performance.