For the universal phenomenon that the plastic film thickness control system had the characteristic of strong nonlinearity, large time delay, low control precision, a plastic film thickness smart control system was designed combined with fuzzy control, neural network and classic PID control. Through combining fuzzy control with neural network, control parameters of PID were regulated online automatically. After field testing, comparing with traditional thickness control, using this system, convergence time is reduced greatly and the control precision is improved, the error of film thickness is reduced to 2 μm.