Abstract: The control and experiment research of the precision positioning system of the grating ruling machine were presented. Considering the non-linearity and time-variation characteristics of the precision positioning system, PID feedback control based on BP neural network was utilized in the positioning system. With the functions of self-learning，the parameters of the PID controller were adjusted in real time, and precision positioning could be realize by the PID controller. Experimental result shows that a positioning accuracy within 5 nm was achieved.

Key words: grating ruling machine, precision positioning, BP neural network, PID control