By designing a SBW (steering-by-wire) system structure of articulated wheel loader and analyzing its steering feel, this paper presented a control strategy on the system, and designed an improved PID (proportional integral derivative) control system based on back propagation neural network. The on-line adjustment of the coefficients of proportion, integral and differential of the classical PID controller was realized by means of the self-study of the neural network, allowing the adaptive control for the steering feel. The simulation and experimental results showed that the back propagation neural network PID could satisfy the requirement of the steering feel.