Research On Multi-Unmanned Vehicle And Multi-UAV Cooperative Transport Path Planning Algorithm

Currently in the field of logistics,many links are automated.However,the "last mile" express transportation is still dominated by labor,and the labor cost it consumes has become a bottleneck in the development of the logistics industry.The experiment of unmanned aerial vehicle delivery has been successful two years ago,indicating that technically the unmanned aerial vehicle can complete the express delivery task.However,how to effectively cooperate with drones and unmanned vehicles to maximize the efficiency of unmanned vehicles and drones is a key issue that needs to be solved urgently.The "last mile" path planning problem is a multi-task multi-unmanned vehicle unmanned aerial vehicle,a collaborative path planning problem with road network restrictions and load restrictions,which is obviously a complex modification of the VRP problem.The VRP problem is an NP-hard problem,so the problem studied in this paper is also an NP-hard problem.For NP-hard problems,although P=NP is still one of the unsolved problems in mathematics,the mainstream idea is that P?NP,that is,an algorithm that does not have a polynomial time complexity can generally find the optimal solution for NP-hard problems.In general,an algorithm to be applied in practice must be polynomial time complexity,so for the NP-hard problem,it is not required to design an algorithm that can find the optimal solution,as long as the algorithm finds a better solution,it can be practically applied.Based on this situation,this paper analyzes the problem model,then designs a heuristic algorithm to solve the path planning problem,and finally tests and summarizes it.The algorithm is divided into two parts: unmanned vehicle planning and unmanned vehicle planning.First,carry out unmanned vehicle planning,and then carry out unmanned vehicle planning based on the results.If unmanned vehicle planning cannot be completed,modify the unmanned vehicle planning based on feedback.Conditions until the unmanned vehicle planning is completed.Unmanned vehicle planning is based on the ant colony algorithm,optimized for practical problems,and added two new parameters.The idea of drone planning is greedy thinking,and path planning is done point by point.The UAV may be unable to complete the transportation for a certain point due to overweight or overdistance.At this time,the reason for completing the transportation of the gameplay is returned to the unmanned vehicle planning,the unmanned vehicle planning is re-run,and then the drone planning is repeated,iterating until the planning is completed.Finally,an experimental platform was built and tested in java language.Theexperimental data comes from the TSPLIB library and randomly generated point data,plus the Delaunay algorithm to generate a triangle network as road network data.Because there are few algorithms for the problem scenes in the text,the experiment is compared with the unoptimized algorithm proposed in this paper,which proves the optimization effect of the algorithm.It also shows that the unmanned vehicle UAV cooperative transportation is much better than the unmanned vehicle alone transportation,which proves the effectiveness and correctness of the algorithm.The algorithm in this paper has played a role in promoting further algorithm optimization in this field,and is helpful for the unmanned landing of the "last mile".