Research on Three Dimensional Trajectory Following Control Method of AUV
|School||Harbin Engineering University|
|Course||Control Theory and Control Engineering|
|Keywords||Autonomous Underwater Vehicles Trajectory Following SDRE Particle Swarm Optimization|
AUV is a complex system with non-linear, coupled and ever-changing model exposed to severe external disturbances, so maneuvering control is a tough issue. In the track-line survey missions, submarine cable detection task, topography, exploration missions, ocean observations, sea level analysis, and military applications require AUV efficiently, accurately tracking a specific curve. In this paper, trajectory tracking control of AUV is studied.Firstly, a 6DOF mathematical model for the AUV was obtained according to Newton’s laws and Rigid-body dynamics. And researched the planar trajectory following strategy for AUV, divided the 6 DOF equations of motion into two non-interacting subsystems, as longitudinal subsystem and lateral subsystem, and design the trajectory following controller based on backstepping strategy, and estimate and compensate the constant environmental disturbance. The simulation results showed that the designed controls perform very well. Backstepping is a design methodology for construction of a feedback control law through a recursive construction of a control Lyapunov function, the design method is simple and flexibility.Secondly, The State-Dependent Riccati Equation (SDRE) strategy is introduced, Which include SDRE nonlinear regulation, existence of solutions, stability analysis and design strategy. SDRE is a nonlinear design method, and without linearization the model. The method entails parameterization of the nonlinear dynamics into the state vector and the product of a matrix valued function that depends on itself and can solved on-line to give the suboptimum control law. Design a three dimensional trajectory controller of AUV based on the SDRE strategy, and simulation results showed that the controller based on SDRE can control AUV following the reference trajectory. We also noticed that the different weighting matrices Q and R displayed distinct performance. So Particle swarm optimization has been used to find optimization Q and R.At the last, The basic particle swarm optimization has been partially improved. The group historical experience particle swarm optimization (GHEPSO) is proposed, particles are not influenced only by the group optimal location of the current iterative time and by their historical optimal location, but also by the group optimal location of previous iterative time at the same time. This algorithm more fully use the group experience information than basic PSO algorithm. The performance of the algorithm is analysed through several typical test functions, camparing this algorithm with basic particle group algorithm. The result shows that GHEPSO is better to solve the problem of multi-modal function than the basic PSO. And the optimized effect will be more improved if GHEPSO, MPSO and TVAC can be combined together. Use the combined PSO search the optimization Q and R. The optimization Q and R showed the better performance for AUV three dimensional trajectory following.