Traditional PID controller parameters are fixed and difficult to meet the shortcomings of the rockets launchers on the requirements of the tracking accuracy and resistance to load capacity, a PID controller based on RBF neural network was designed. Firstly, by improving the network of dynamic resource allocation algorithm, neural network structure was completed, and then the neural network was simplified to be successfully applied to real-time adjustment of the PID control parameters in the DSP processor. The experimental results show that this control strategy can effectively improve the tracking precision and resistance to load capacity.