Research On Mechanism And Trajectory Optimization For Dynamic Soaring With Fixed-wing Unmanned Aerial Vehicles

In military operations,unmanned aerial vehicles(UAVs)perform missions of intelligence,surveillance,and reconnaissance(ISR)and long-range precision strike.The application of UAVs is considered as the core military capability and the effective means to win the local war under the condition of informatization in the future.UAVs are also used for civil missions,such as geological mapping,resource survey,environment monitoring,meteorological observation,disaster assessment and rescue and so on.These missions are either difficult or hazardous for human and manned aircraft and usually illustrated by three Ds:?dull?,?dirty?and?dangerous?.Looking from the history perspective,application background and current research directions for UAVs,improving the long-endurance performace to extend the duration and coverage area in task performing is the main development tendency.From the viewpoint of concept design for UAVs,the requirement of endurance is the springboard and goal,and the endurance performance is finally determined by onboard energy consumption.However,UAVs suffer from onboard energy storage limited in size and weight,thus the innovative approachs should be explored to improve the endurance performace in terms of energy.Albatrosses use a flight technique called dynamic soaring to gain energy from wind gradients near the ocean surface to travel for a very long journey and period,even around the world,almost without flapping their wings.They have a unique?shoulder lock system?that keeps their wings outstretched without any muscle activity expending energy.Thus they can be treated as fixed-wing UAVs.In addition,wind gradients are persistently distributed in different altitudes with various dimensions.Fixed-wing UAVs therefore have great potential to use dynamic soaring extracting energy from the environmental wind field to achieve long-endurance and broad-coverage flight,even to realize forever enginless airborn.However,so far,the energy-gain mechanism in dynamic soaring is not explicit,and additionally,for the way,means and allowable wind conditions for dynamic soaring,further studies are expected.The subject investigated of the thesis is dynamic soaring with fixed-wing UAVs,and the detailed work is sumerized as following:(1)Dynamic models are derived and trajectory optimization method is proposed for UAVs in dynamic soaring.Two different dynamic models(equations of motion,EOM)of a UAV have been derived:one is expressed in body-fixed frame and the other in air-relative flight path frame.These two models are equivalent to each other but each one has appropriate applicability.The former is used in flight simulation,while the latter is used for energy analysis and trajectory optimization.An efficient direct collection approach based on the Runge-Kutta integrator is proposed to solve trajectory optimization problems for overall dynamic soaring patterns.(2)A guidance and control strategy based on Rayleigh cycle is desined to perform dynamis soaring.Rayleigh cycle includes sequentially upwind climb,downwind turn,downwind dive and upwind turn.The optimal climb angle for maximizing energy harvesting during climbing and diving and the optimal roll angle for minimizing the nenergy loss are computed according to equations of motion.A guidance and control strategy is desined to tracking the optimized Euler angles to realize the energy-gain Rayleigh-cycle-like dynamic soaring.The energy gain then is released in a level flight at low atltitudes toward a specified travelling direction,thus the UAV has the ability to follow asymptotically the large-scale straight-line or sinuous path.A method of wind parameter estimation based on the augmented extened Kalman filter is proposed,and validated with the simulation results of Rayleigh-cycle-like dynamic soaring as simulative sensor measurements.(3)A comprehensive explanation for dynamic soaring mechanisms with respect to different frames of reference is provided.By discussing the albatross's EOM analytically and the numerical solution of optimal dynamic soaring,energy-gain mechanism is explained and compared with analyses in previous publications.The analysis indicate that the energy gain in the air-relative frame originates from alternately crossing the large wind gradients,while the energy gain in the inertial frame comes from lift vector inclined to the wind direction.These two concepts are not equivalent in terms of energy harvesting but should be realized simultaneously in the energy-neutral dynamic soaring cycles.The wind gradient has proven to be the essential energy source to sustain the energy-neutral cycle in both frames rather than the wind speeds considered in some previous studies.(4)A complete categorization system for optimal dynamic soaring trajectories is proposed.According to four essential phases in the Rayleigh cycle and the albatross's real flight pattern,different dynamic soaring patterns are schematically depicted.The flight trajectories for these patterns are computed by the proposed optimization method combined with the initial guesses and boundary conditions referring to the schematic diagrams.The optimal dynamic soaring trajectories are classified into two closed patterns including?O?shape and?8?shape,and four travelling patterns including???shape,???shape,?C?shape and?S?shape.The characteristics for each pattern and correlation among these patterns are analysed and discussed.The completeness of the categorization is confirmed by listing and summarising dynamic soaring trajectories shown in studies over the past decades.Various patterms indicate that UAVs can simulate albatrosses and even surpass the bird to dynamic saring in more diversified manners.(5)Allowable domain of wind condition for dynamic soaring is expanded by considering the power law exponent(p).Based on the results of trajectory optimizations,the allowable wind condition which involve the permissible power law exponents and feasible reference wind speeds(V_R)supporting dynamic soaring are proposed for both the closed pattern and travelling pattern.The allowable wind condition is exponded from the single wind speed range to a more general two-dimension domain about V_R and p.Sensitivity of the allowable domain to vehicle model parameters,minimum safe height and travelling speed requirement is discussed.(6)The trajectory optimization for using both the potential energy and the wind gradient with a solar-powered UAV is studied.While the vehicle alternately climbs and dives to respectively store and release the potential energy,the dynamic soring mechanism that the vehicle should climb upwind and dive downwind across a wind gradient profile alternately is introduced.The trajectory for combined use of the potential energy and the wind gradient is optimized.The results show that the vehicle extracts more energy from the environment,therefore the necessary onboard energy storage can be diminished and the energy-neutral performance within 24 hours can be improved.This thesis investigates the energy-gain mechanism in dynamic soaring in depth,preliminarily solves the problem of closing the loop of dynamic soaring including wind field estimation,nominal motion planning and low-level control,analyses the possible patterns and the allowable wind condition for dynamic soaring according to trajectory optimizations.This thesis proposes an innovative idea of utilizing the environmental energy sources with fixed-wing UAVs,contributes to the field of dynamic soaring by laying a solid theorietical basis in energy extraction from the wind gradient,and provides a significant guidance to improve the long-endurance performance in UAV applications.