Research On Trajectory Optimization Of Unmanned Aerial-aquatic Vehicle

Unmanned aerial-aquatic vehicle(UAAV)is a new type of aircraft designed by combining the motion characteristics of unmanned aerial vehicle(UAV)and unmanned underwater vehicle(UUA).It can navigate both in air and underwater.Considering that the diving motion of UAAV plays an important role in the performance of UAAV and is being explored,the dynamics modeling and trajectory optimization problem are studied in this thesis.Firstly,the trajectory optimization problem of UAAV is studied.The trajectory optimization problem of two typical aircrafts,folding wing and four-axle eight-rotor UAAV,is studied.The biologically inspired transition strategies for direct diving and indirect diving are designed,and the applicability of the two strategies is analyzed.In order to optimize the diving trajectory under the two diving strategies,the idea of updating the next generation particles with the global optimal solution combined with the particle swarm optimization(PSO)algorithm was improved to improve the teach-learn based optimization(TLBO)algorithm.An improved teach-learn based optimization(ITLBO)with faster convergence speed and stronger local search ability is proposed to optimize the diving trajectory.Then the dynamic models of two typical aircrafts are established and the trajectory control strategy is designed.When establishing the folding UAAV dynamic model,the complicated fluid forces are divided into ideal fluid forces and viscous fluid forces for calculation.The folding wing UAAV adopts a direct diving strategy.The diving process is considered as free movement to avoid the instability caused by switching between different control strategies between air and water.Therefore,the diving trajectory of the folding UAAV depends on the initial state before entering the underwater.In establishing the dynamics model of the four-axis eight-rotor UAAV,the motion characteristics of the four-axis eight-rotor UAAV itself and the gradual change of fluid force,additional mass and buoyancy when diving are fully considered.The four-axis eight-rotor UAAV adopts a indirect diving strategy,and a hybrid control system is designed to control the motion trajectory of the four-axis eight-rotor UAAV.The classic PID control algorithm is designed to control the motion of the four-axis and eight-rotor UAAV in the air and underwater.The PSO algorithm is used to optimize the rotor speed in the vicinity of the water surface to achieve accurate control of the diving process.Finally,in order to obtain a satisfactory diving trajectory under certain task indicators,the PSO,TLBO,and ITLBO algorithms were used to optimize the initial state of the folding wing UAAV before diving and the rotor speed of the four-axle eight-rotor UAAV,and Compare and analyze different algorithms.The simulation results show that the established UAAV dynamic model is reasonable and can reflect the characteristics of diving movement.The proposed ITLBO algorithms is superior to PSO and TLBO in terms of convergence speed and local search ability when optimizing diving trajectory.Finally,different optimization algorithms are selected to optimize the diving trajectory according to the task requirements.