| With the development of UAV technology, UAVs are increasingly being applied to many fields, and tasks must be allocated before UAV executing missions, so researches of multi-UAV cooperative task allocation problem show great significance. Exact methods can solve the problem with optimal solution, but it often takes too long, while intelligent methods can obtain satisfactory solution in a short period of time, but for a large problem, it is still difficult to shorten the solution time in seconds. Changing battlefield environment requires algorithms to solve task allocation problems with a higher speed. The solution speed should be improved continually in order to ensure the timeliness of the results of task allocation.For multi-target assignment problems under the scenario of UAV teams attacking ground targets, this dissertation carry on the research from two aspects:(1) Research on basic models and algorithms for multi-UAV target assignment problems. After analyzed and summarized, the problems are divided into one-to-one, one-to-many, many-to-many issues according to the allocation relationship between the UAVs and the targets, and the basic models of the issues are given. Finally, some commonly used algorithms for solving target assignment problem are summarized.(2) Research on the framework and model of two-stage method in solving cooperative target allocation problems in UAV teams. After the problems analyzed, a two-stage method framework, consisting of assignment at team level and assignment within teams, is proposed to solve the problems. The objectives and models of two stages are analyzed separately. The framework decomposes a cooperative target allocation problem in UAV teams into several sub cooperative target assignment problems within teams, so the sub problems can be solved concurrently, which effectively improves the solving speed of the whole problem.Finally, simulations are performed, and the results show that the two-phase method is feasible and effective. |
PDF Full Text Request