Research On Autonomous Underwater Vehicles Formation Control And Visual Simulation

In the 21st century, governments worldwide asserting more devotion over the marine,which established blue economy industrial chain, unprecedented wealth of material conditions were offered to the research of the marine. Among those studies, Autonomous Underwater Vehicles (AUV) has many advantages such as small size, light weight, low cost and strong independence, relying on those advantages, AUV is now widely used in deep sea reconnaissance,searching and rescue and military affaire. AUV's control system is the key of the whole AUV technique, so it is necessary to study the control system, algorithm and analog simulation. Compared with single AUV, multi-AUVs system is able to bear more complex and diversity mission, and thereby the control system of multi-AUVs is more complex, which contain both single AUV control system and coordination control part.AUV formation control and the visual simulation relevant are studied in this paper. From those two aspects, the AUVs formation control, path planning and obstacle avoidance are extensively studied. The application of a new type 3D development platform in AUV's visual simulation is explored, with the purpose of enhance the dynamic quality of the AUV simulation, so the verification of the algorithm become more clear and intuitive. The main works are as follows:Firstly,AUV formation control system base on six-degree of freedom AUV is established with leader-follower formation control strategy. According to the mathematical model of AUV and requirement of the distance between the AUVs, error model is build. And then the controller of follow AUVs is derived based on the Lyapunov stability theory.Secondly, the fast-marching algorithm is used to realize the path planning of the leader AUV in the formation. Optimal function (distance function) is build for each grid from the grid space, then through solving each grid's optimal function value, a non-collision path is produced.Thirdly, the biological inspired neural network method is used to enhance the capability of obstacle avoidance for the follower AUV in the formation. Grid model of the all underwater environment is established for algorithm, each grid is defined as a neuron, an safe path can be calculated by comparing the value of the neurons' active. Obstacle avoidanceFinally, according the overall framework and demand analysis of the simulation system,Unity 3D and 3DMax are used to build the system. AUV's 3D model is build by 3DMax, and exported as .FBX file format which is compatible with Unity 3D. Using the script system of Unity 3D to realize the algorithms, and tiling them to the targets in the 3D viewing, then the whole visual simulation system is complete.