Research On UUV Real-time Path-planning Before Recovery By Stinger-principle

UUV recovery technique is an important technical support for UUV completingvarious missions. The stinger-principle recovery technique is an effective methodappropriate for complicated oceanic environment, with UUV navigating to the stern of themother ship with a certain heading to recovery from surface point. Owing to navigationerror, the random surface point position and mother ship drifted motion with environmentmay induce the difficulty for path determinant. Thus, this thesis develops the research onreal-time path planning of stinger-principle recovery for UUV.By analyzing the process of the stinger-principle recovery, the problem of the pathplanning is divided into three parts: global path planning under the influence of oceancurrent, local path planning in dynamic obstacle zone and the path planning for the stern.Considering these problems, a real-time path planning method is proposed for recovery.The main contents in this paper are shown as following:Firstly, considering global path planning under current influence, a heuristic A*algorithm is used to design global planning path with multiple constraints, and Beziercurve theory is applied to smooth global path for the seaworthiness. Simulationexperiments are performed to illustrate the feasibility of the algorithm in the steady flowand vortex environment.Secondly, for local path planning in dynamic obstacles area, the dynamic potentialfield is improved to plan path, taking account of the current disturbance and the resultantforce on UUV. To evaluate the effectiveness of the improved algorithm, simulations in thesteady flow and vortex environment are designed.Thirdly, to satisfy the demand of heading at the end of the stern-planning path, aRecovery Path Terminal Guidance approach is proposed. According to UUV characteristic,a semi-circle path is set at the stern with a target point on it guiding UUV to navigatealong this known path. By considering different heading conditions, Simulations aredesigned to evaluate the performance of the proposed approach.Finally, synthesizing path planning methods above all, a real-time path planningalgorithm is designed for the stinger-principle recovery. The effectiveness of the algorithmis validated by different cases. Simulation results demonstrate that the real-time pathplanning algorithm proposed in this paper is efficiently to plan path and can handle therandom watered points caused by the navigation error.