Research On Multi-uav Track Planning Based On Collaborative Emergency Rescue

In recent years,frequent disasters have caused heavy damage to human life and property.Due to difficulties in obtaining information about the disasters,emergency rescue is difficult to be operated accurately,leading to further destructive disasters.According to its preponderant characteristics,drone can be used for disaster relief and guidance in rescue missions.The route planning methods of drone are required urgently to complete the investigation and rescue missions and coordinate with other existing drones.In real situations,such as earthquakes,reports show that in some areas there are probably life-source signals existing,which need to be checked first by drones.In other areas without report of life signals,drones need to perform the covering mission to explore life sources inch by inch.Similarly,the investigation of the sources of disasters is also true.Combined with the practical situations,prioritized detection of known target points and covering mission of unknown target points were fully researched in this paper,by the process of analysis,modeling,algorithm improvement and simulation experiment.Models can be builted based on the UAV model,collaborative self-organizing system and route planning targets.By analyzing the feasibilities of intelligent algorithms such as ant colony algorithm,genetic algorithm,neural network algorithm,simulated annealing algorithm,voronoi graph algorithms,etc.,ant colony algorithm was chosed to solve this problem finally.The first problem can be solved base on the Hamiltonian path by using virtual node and shadow ant approach.It is necessary to complete the traversal detection for all known targets.The target start point and the target end point do not need to be coincident.This is similar to the TSP(Traveling Saleman Problem).Lingo and Matlab software are used respectively to perform computational verification and simulation experiments in different numbers of drone scenarios.Simulation and experiment results prove its feasibility.In the second case,the design of the scheme and the algorithm for the cooperative mission of the UAV are performed for the area coverage target.A strategy for dividing regions based on UAV performance index is proposed.The first division of the area is based on the minimum number of turns,the shortest flight width,as the cutting area basis.Then the sub-area is detecting by scanline-based coverage.Eventually,the task of rapid collaborate investigation of the target area was complete in practical cases.