Motion Planning And Stabilization Of Multi-Vehicle Formation

Motion planning and cooperative control of multiple autonomous underwatervehicles (AUVs) is important in the construction of dynamic adaptive oceanprediction networks. Multi-AUV formation is effective in large scale oceanobservation missions. Individual vehicle in the formation simultaneously samples inthe observation area. The sampling data obtained by the formation can be used tomodel physical processes in the ocean with a higher accuracy than those developedby sampling data from a single vehicle. This dissertation presents motion planningand stability analysis for multi-AUV formation in complicated ocean environment.The main contributions are summarized as follows:The multi-vehicle formation is regarded as a multi-body system and theindividual in the formation is treated as one rigid body in the multibody system.Artificial potential fields that include interactive potential and orientation potentialare introduced as functions of relative position and relative orientation betweenneighboring vehicles respectively. Virtual forces derived from the constructedartificial potential fields are used to constrain the motion of individual vehicles.Motion planning for the formation is conducted using the artificial potentialfield (APF) method and multibody dynamics. AUVs in the formation are treated asparticles with full actuation. The Multi-AUV formation is simplified as onemulti-particle formation by ignoring the attitude of the vehicles. The leader-followerapproach is used to coordinate individuals in the formation and Kane’s method fordynamic analysis of multibody systems is used to study the dynamic characteristicsof motion of the multi-AUV formation.The dynamic model of the multi-rigid-body formation is constructed based onAPF method and solved using multibody dynamics. The multi-AUV formation issimplified as a formation of spinning rigid bodies by ignoring translation of theAUVs. Coordination of the multi-spinning-body formation is investigated. Theattitude synchronization of the spinning rigid body formation implies that theindividuals in the formation have the same rotation matrices. Simulation is alsoperformed on coordination of the multi-AUV formation as a group of spinning rigid bodies.The stabilization control law for the multi-AUV formation is designed based onenergy shaping. A new energy function that keeps Lagrangian structure of the systemis introduced by feedback to the formation and a potential shaping function ofrelative orientations is introduced to synchronize the motion of the vehicles. Theenergy method is used to prove stability and coordinated behavior of the entireformation.