Abstract: A neural network adaptive inverse control strategy was proposed to control the distortion of parametric array system which is a strong nonlinear acoustic system. The inverse model of parametric array system was introduced. A pseudo-line system was generated from the parametric array model and it's inverse model. Three parameters of PID were adaptively controled by the BP network and the system was controled by PID. Simulation research was carried out by establishing the combined inverse control model with MATLAB. Simulation results showed that the neural network PID control had good accuracy and adaptability such that good control effect could be achieved.