Finite-time Control For Autonomous Underwater Vehicle

With the rising growth of population and the increasing shortage of landresources in the world today, the ocean which obtains much more abundant resourcesis becoming the strategy center of the global economy. As a country which has hugepopulation, China, especially, can take the advantage of its great ocean resources toquicken its economic development. the first underwater robot has been invented inorder to deal with the severe natural conditions of the ocean. And nowadays, as onekind of the underwater robots, the autonomous underwater vehicle has become animportant research subject of ocean science and military. In order to improve thecontrol accuracy and the disturbance-rejection properties near the equilibrium pointof the autonomous underwater vehicles, finite time control laws are used to designcontrol algorithms.The finite time control problem for the course angle of the autonomousunderwater vehicles is discussed. The main results are as follows:Firstly, the research status and the main control methods for the AUV aresummarized. Both the modelling and modelling simplification for the course angle ofthe autonomous underwater vehicles are introduced. The the definitions and thesufficient conditions for finite time stability are given.Secondly, the constant-speed converging terminl sliding mode control algorithm,fast terminl sliding mode control algorithm and fractional power order terminalsliding mode control algorithm are designed for the simplified mathematical modelof the autonomous underwater vehicles. Their characteristic are analyzed andcompared based on the the simulation results.Finally, a class of finite time PID-like control algorithm is designed for thesecond order affine nonlinear control systems and its finite time stability are provedbased on the homogeneity. This class of finite time PID-like control algoritm is usedto control the course angle of the autonomous underwater vehicles. Its efficiency andbetter robustness properties for modelling errors are verified by the simulationsresults.