Aiming at heavy calculation and low robustness and response hysteresis to strong disturbance of fuzzy neural network controller for autonomous underwater vehicles,T-S fuzzy neural network control based on hybrid learning algorithm was proposed.The parameters of the membership function are opti-mized by immune genetic algorithm off line and neural network on line to reduce the calculation of neural network and enhance the response ability to environment changing.Moreover,T-S model is used to adjust the dynam...