The natural cooling after the rolling of the 100-meter-long rail tends to curve the rail from the bottom to the head. Based on the analysis of the rail pre-bending process, this thesis proposes the implementation plan of the pre-bending vehicle control system. The position loop uses fuzzy neural network regulator and the speed loop uses PB-based neural network PID regulator. Also the thesis gives the corresponding control algorithm. The control equipment uses Siemens advanced SIMATIC PCS7 process control system. Through experimental analysis, the dynamic tracking performance and the static accuracy of the system are guaranteed and the vehicle is accurately placed as expected.