The work aims to improve the packing precision of quantitative weighing packaging machine. A quantitative weighing packaging control system was designed with the programmable logic controller (PLC) based on traditional packaging machines. Combining RBF neural network and PID control method, a quantitative weighing packaging control strategy was proposed in order to realize the closed-loop control. The proportion, integral and differential coefficient of PID control were adjusted rapidly with the RBF neural network self-learning ability. The control system hardware structure was designed with PLC as the core and the composition of human-machine interface was presented. A lot of experiments showed that the packaging precision was greatly improved and the maximum relative error was no more than 0.6%. The method described herein has enhanced the precision, reliability and maneuverability of quantitative weighing packaging machines.