Guo Hongda

The improved adaptive hybrid fuzzy control of AUV horizontal motion

An improved adaptive hybrid fuzzy control algorithm was put forward which was based on Mamdani inference, due to the steady-state error, the insensitive control output in the region near zero, the lack of adaptive capability and other issues of traditional fuzzy control. And the corresponding controller was designed. Weights was adjusted by changing adaptive parameters online, so that we could control AUVs both in the initial phase and the steady-state phase. It didnt only improve the response speed of control, but also reduce the oscillation magnitude of static phase. The self-tuning performance of the controller was extremely improved. To test the performance of the controller, the horizontal dynamic model of an AUV was established which was designed by Science and Technology on Autonomous Underwater Vehicle Laboratory, Harbin Engineering University. The decoupling control was executed to control the longitudinal speed u and heading angle Ψ separately. Simulation results showed that, in the presence of sensor errors and external environmental interference, the improved adaptive hybrid fuzzy controller had a superior control performance.