Research On Uav Path Planning And Scheduling In The Air-to-ground Collaboration Scenario

Data collection of disaster area plays a very important role in emergency scenarios,which can support emergency applications such as target identification,data analysis and high-definition map reconstruction.However,due to the damage of buildings in the disaster area and poor roads,some sensor information deployed on the ground can’t be collected in time,which affects rescue efficiency.Therefore,unmanned aerial vehicles(UAVs)can be used to assist data collection with sensors deployed in emergency scenarios.Nevertheless,due to the constraints of the UAV’s energy,its flight range and duration are limited.It is necessary to optimize the UAV’s path to maximize the efficiency of information collection.In this context,this article focuses on the path planning of UAVs in emergency scenarios during data collection.The path planning of UAV is studied in this paper.As a result of the research,a system for path planning and video-capturing under the scenario of a single UAV without charging is built.First of all,in order to improve the efficiency of data collection and shorten the time of data collection,this paper analyzes the collaborative data collection scenario of UAVs in the emergency scenario,and studies its path planning and task assignment respectively.Firstly,in the aspect of path planning,considering that the UAV may not be able to collect all the sensors due to its limited energy,this paper aims to maximize the weighted sum of geographical fairness and data collection amount,and proposes genetic algorithm and dynamic programming algorithm to achieve low complexity and high precision path planning results respectively.Then based on the results of path planning for UAV,considering the different utility difference problem of allocation,this paper considers the geographical fairness,data acquisition,UAV’s energy consumption and so on.The regional preference list is established by the path planning results,and Gale-Shapley algorithm will realize the stable one-to-one match between the UAVs and sub-regions.The simulation results show that the proposed path planning scheme takes into account both geographic fairness and the amount of data collected under the UAV energy constraints,and significantly reduces the path length of the UAV,while Gale-Shapley algorithm with a low complexity maintains the performance close to the optimal allocation.Then,for reducing the cost of UAVs and the ability of UAV to be recharged and reused for several times,this paper then analyzes the rechargeable scenario of a single UAV,and studies the path strategy and charge strategy of UAV.Considering the single UAV does not cover the entire collection area because of its finite energy.Through sensor clustering and cluster head selection,UAV can only collect cluster head nodes.Based on this basis,this paper proposes a path planning scheme based on deep reinforcement learning,in order to realize the UAV’s trajectory and the adaptive adjustment of the number of charging and complete the joint optimization of flight strategy and charging strategy.The simulation results show that the proposed algorithm greatly reduces the energy consumption of UAV,which illustrates the rationality and effectiveness of the proposed scheme.Finally,in order to verify the theoretical path planning algorithm proposed in this paper,this paper further builds a path planning and data collection system for a single UAV above path planning scheme design and tests it in the actual scene.To be specific,the system controls the path planning of the UAV by mobile phone and collects data using sensing equipment such as camera carried by the UAV,and sends it back to the ground in real time.Through the actual test,this paper verified that the system can effectively reduce the length of UAV flight path,improve the efficiency of UAV data collection,and has a certain practical application value.