Multiple Models Switching Control For Unmanned Underwater Vehicle

Unmanned underwater vehicle(Unmanned Underwater Vehicle,UUV)works in complex environment.When the fluid density,work load carrier or performing different tasks change,it usually leads to changes in working conditions.However the traditional single model control is difficult to meet dynamic performance in the deviation operating conditions,for example,the PID control method,its drawbacks can not be neglected such as large overshoot or excessive differentiation element lead to instability phenomenon in system.So we propose a method to build the model set and controller set in all conditions.We call this method multiple model control.The major contents are organized as follows: building of the model set,selection of the controller algorithm,selection of switching strategies,testify of the stability.Firstly we analyze the UUV model,complete the model set according to its dynamic model and kinematic model,and decoupling UUV mathematical models into velocity model,horizontal model,vertical model.This paper decided to use longitudinal velocity conditions as the dividing point for local model to build model set.Then we can construct horizontal model set and vertical model set in different speed.To ensure the optimal control performance of state variables on each degree of freedom,we must design the best control algorithm of fixed model.For example,a fixed model in longitudinal speed of 2m/s,we used PID,sliding mode,fuzzy control algorithms for UUV longitudinal speed,heading,depth control.To choose appropriate control algorithm as the best control algorithm of fixed model according to corresponding analysis.Then we use this control algorithm to control model set and complete the corresponding design of controller in different working conditions.There are two ways to achieve switch in controllers of fixed model set T-S fuzzy switching(soft switching)and indirect performance indicators switching(hard switching).For example,in horizontal plane,we compared tracking control performance of heading control with multi-model switching control and single model control in current disturbance case.We can find that multi-model switching control can improve the dynamic performance of the control system in changing operating conditions.Then we contrast hard and soft switching mode to point out their strengths and weaknesses,and propose a way to prove the stability of soft switching.Finally,we applied multi-model switching control in path tracking of UUV and used the line of sight guidance as a method to control heading.In changing operating conditions,we use multi-model switching control algorithm to track the heading angle and the pitch angle.Respectively,we contrast results of path tracking of UUV to validate the effectiveness of multi-model switching control algorithm.