In order to solve the problems that the system parameters of the deep electro-hydraulic servo system are difficult to determine, the internal parameters have time-varying and the external load disturbance is large, a control strategy combining the PID controller and the neural network is designed. Analyze the mathematical model of the deep electro-hydraulic servo system and the structure and working principle of the controller. The radical basis function neural network is used to dynamically modify the control parameters of the PID controller. The particle swarm algorithm is used to select the optimal neural network right offline. Use Matlab in controller of certain deep electro-hydraulic servo system, and compare it with the classic PID controller and RBF-PID controller. The simulation results show that the controller has good fast response capability and robustness.