Rugged Seabed Observation Oriented Three-dimensional Path Planning Of An Autonomous Underwater Vehicle

With the rapid development of underwater equipment technology,Autonomous Underwater Vehicle(AUV)has shown broad application prospects especially in ocean observation.Among them,the autonomous path planning technology is the fundamental guarantee to ensure the optimization and safety of the AUV’s arrival in the observation area.This thesis is oriented to the needs of rugged seabed observations,and proposes a threedimensional AUV path planning scheme that combines path optimization and navigation safety to solve the problem of AUV global-local path planning and AUV area coverage path planning in rugged seabed environments.The main research work is as follows:Firstly,to meet the needs of seabed observation and aiming at the safety of AUV navigation in rugged seabed environment,a rugged puffing method for seabed area is proposed.Specifically,the terrain state space is obtained through grid modeling of the rugged seabed environment,and the area is decomposed;and then the ruggedness of each area is defined to expand at different scales,thereby establishing a path planning that takes into account the safety of AUV navigation and rugged seabed observation environmental model.By using the popular A* algorithm to conduct path planning analysis in the terrain environment before and after the expansion,the effectiveness and superiority of the proposed rugged expansion method are fully verified.Secondly,for the global path planning of AUV in rugged seabed environment,smooth Lazy Theta* and improved elite genetic algorithm are proposed respectively.Specifically,the smooth Lazy Theta* algorithm overcomes the drawbacks of the Lazy Theta* algorithm in terms of path length;The improved elite genetic algorithm improves the search efficiency of the elite genetic algorithm in complex spaces,and designs a new initialization method.A large number of simulation verification and comparative analysis show that the smooth Lazy Theta*algorithm can obtain a better global path in the optimal time,and the improved elite genetic algorithm can obtain the optimal length of the global path.Thirdly,considering the local path planning of single AUV and multiple AUVs under AUV dynamic constraints,an improved artificial potential field method and priority strategy are proposed.Specifically,the improved artificial potential field method fully considers the AUV’s bowing,pitch angular velocity,and forward velocity constraints,and by improving the repulsion potential field function and combining the smooth Lazy Theta* algorithm,solve the problem of unreachable target and local minimum in traditional artificial potential field method;Different priorities are defined for multiple AUVs,which realizes the collaborative planning among multiple AUVs,and ensures the safety of collision avoidance in local path planning of multiple AUVs.The simulation results verify that the improved artificial potential field method can solve the local minimum problem and the target unreachable problem,and the improved artificial potential field method combined with the priority strategy can realize the local path planning of multiple AUVs.Finally,for single AUV and multi-AUV area coverage path planning,a coverage elite genetic algorithm is proposed.Specifically,combined with the rugged puffed seabed environment model,the puffed grid points are used as the points to be covered,and then based on the iterative idea of the elite genetic algorithm,the overriding elite genetic algorithm was proposed by redesigning the genetic operator.It realizes the traversal coverage of the observation point of the observation area by the AUV,while ensuring the safety and optimality of the observation path.The simulation results verify that the coverage elite genetic algorithm can obtain the optimal area coverage path for single AUV and multiple AUVs.