Abstract: An adaptive neural network fuzzy controller for pneumatic servo system based on electric-pneumatic pressure proportional control valve is studied. By trained off-line the neural network Identifier (NNI) is able to approach the model of nonlinear dynamic system and regulate the control rules of the fuzzy controller (FC) on-line. The position control precision and servo characteristic is increased. The experimental results show that the proposed controller has excellent performance for the pneumatic servo system and has the ability of self-adaptive.

Key words: Adaptive control, Fuzzy control, Neural network identify, Nonlinear dynamic system, Pneumatic servo system