Research On Ground-air Integrated Mobile Edge Network For Smart Grid

At present,the problems of environmental pollution,low energy efficiency,and high external dependence brought about by the fossil fuel-dominated energy structure are still serious in China.In order to ensure the sustainable development of economy and society,China must gradually get rid of its excessive dependence on fossil energy,and the reform of energy structure is imperative.Relying on policy support and technological upgrading,clean energy has shown a trend of rapid development,and the proportion in the energy structure has increased steadily.In the implementation process,the economic operation of clean energy such as wind power and the stable operation of power transmission and transformation network are particularly important.This thesis studies the ground air integrated mobile edge network for smart grid,aiming at improving the detection efficiency of wind turbines,the communication network recovery ability in extreme cases,and the communication ability in remote areas,which has practical significance for the development of clean energy.Relying on unmanned aerial vehicle(UAV)edge computing,this thesis focuses on two representative application scenarios: wind farm routine inspection and post-disaster area communication emergency response.Based on the different requirements of task detection,data communication,computation offloading and energy supply in these application scenarios,the corresponding ground air integrated mobile edge network are proposed.Then,the trajectory planning problem,the computation offloading optimization problem and the robust optimization problem under location uncertainty are discussed and analyzed.The main contents of this thesis are as follows.1.The problem of wind farm routine inspection based on UAV edge computing is studied.In this scenario,the UAV detects multiple wind turbines and processes the detection data through local computation or computation offloading.Considering the influence of wind speed and wind direction on UAV flight energy consumption,two optimized UAV flight speeds method are proposed according to the given wind speed,and a detection trajectory planning scheme is given according to the change of wind direction.Meanwhile,considering the coupling among communication trajectory and computation offloading,an iterative optimization algorithm is proposed to minimize the energy consumption of data processing.2.The problem of wind farm routine inspection based on space-air-ground integrated edge computing is studied.In this scenario,the UAV not only detects wind turbines in multiple sorties,but also processes the detection data in an appropriate way of computation offloading.While,the UAV can offload the sensory data to the ground station or low earth orbit satellite optimally.In order to overcome the influence of wind on UAV trajectory planning,a low complexity wind turbine inspection trajectory planning and UAV scheduling approach is proposed firstly.Then,combined with the different characteristics of satellite network,air network and ground network,the communication trajectory and computation offloading under different communication modes are jointly optimized.3.The problem of network offloading in post-disaster area based on UAV edge computing is studied.In this scenario,the UAV provides charging and computing services for ground devices in the disaster area.Considering the influence of the location uncertainty of devices on UAV trajectory planning,a joint resource allocation and trajectory planning algorithm for given devices location is proposed firstly.Then,the cut-set method is adopted,by updating the worst case device locations under different UAV flight trajectories,to achieve the purpose of eliminating the uncertainty of equipment position.4.The problem of resource allocation for UAV edge network based on device random distribution is studied.In order to further improve the energy acquisition efficiency of ground device,the research is carried out from two aspects: reducing the communication energy consumption of device and reducing the pessimistic estimation of device location.Firstly,Bernstein type inequality is used to eliminate the influence caused by the probability distribution of device location,and the successive convex approximation method is used to optimize the UAV trajectory.Then,combined with the time division multiple access communication mode,a joint device communication scheduling and UAV trajectory planning algorithm is proposed.