Title :
Design of inverse controller of zero-speed fin stabilizer based on RBF neural network
Author :
Song Jiguang ; Liang Lihua ; Zhang Songtao ; Qi Zhigang
Author_Institution :
Coll. of Autom., Harbin Eng. Univ., Harbin, China
Abstract :
The hydrodynamic of zero-speed fin keeps no longer approximate ratio relationship with fin angle, but has a memory nonlinear mapping with the fin angle, angular velocity and angular acceleration, besides its nonlinearity and uncertainty of ship roll motion equation. It not only caused the traditional adversarial based PID controller can´t get operation variable directly, but also severely reduced the effects of anti-rolling about different periods of interference and high sea state. A separating strategy was applied in order to simplify the system structure according to system input/output nonlinear of zero-speed fin stabilizer, and a master-slave controller based on RBF neural network adaptive inverse control and GRNN approach mapping between fin angle and force. The simulation results show that this method can improve the effectiveness of the anti-rolling at different periods of interference and high sea state, and have self-adaptability with the varying of ship roll motion equation.
Keywords :
acceleration control; adaptive control; angular velocity control; control nonlinearities; control system synthesis; marine control; motion control; neurocontrollers; nonlinear control systems; radial basis function networks; self-adjusting systems; ships; three-term control; GRNN approach; PID controller; RBF neural network adaptive inverse control; angular acceleration; angular velocity; antirolling; fin angle; fin force; high sea state; hydrodynamic; interference; inverse controller design; master-slave controller; memory nonlinear mapping; nonlinearity; self-adaptability; separating strategy; ship roll motion equation; system input/output nonlinear; system structure; uncertainty; zero-speed fin stabilizer; Educational institutions; Electronic mail; Equations; Hydrodynamics; Marine vehicles; Mathematical model; Neural networks; Fin stabilizer; Inverse control; RBF network; Zero speed;
Conference_Titel :
Control Conference (CCC), 2014 33rd Chinese
Conference_Location :
Nanjing
DOI :
10.1109/ChiCC.2014.6896957