Task Assignment Strategy Of Heterogeneous Multi-AUV System Based On Auction Mechanism

The task assignment problem of Autonomous Underwater Vehicle(AUV)system is a complex non-deterministic polynomial optimization problem.It is difficult to find the global optimal solution of the problem.Especially in the complex and changeable underwater environment,the multi-AUV system task assignment faces multiple challenges,such as the limited energy resources,low overall efficiency of the multi-AUV system,and the uncertain influence of underwater ocean currents.In view of the above issues,a reasonable and effective task allocation algorithm is proposed in the thesis to improve the efficiency of the multi-AUV system and increase the ability of AUVs in dealing with the complex underwater environment.First,the AUV types,task types,and the architecture of multiple AUV systems are introduced in the thesis briefly.Then,Based on the analysis of task scenarios,AUV system mobility and task uniqueness,a heterogeneous multi-AUV system task assignment optimization model with multiple constraints is constructed.In addition,The influence of underwater currents on task assignment of heterogeneous multi-AUV system is analyzed.Secondly,a distributed robust auction task assignment algorithm is proposed to solve the problem of task assignment under the conditions of energy constraints and uncertain underwater ocean currents.The existing auction algorithms ignore the interests of auctioneers and do not conform to the law of the market.Therefore,the task reward feedback mechanism is introduced.This mechanism not only ensures the utility of AUV system,but also reduces the cost of task distribution system(auctioneer).In addition,the proposed algorithm employs robust theory to optimize the uncertainties generated by the task assignment model of ocean currents.This theory enables the multi-AUV system to maintain preferable utility in uncertain environment,and improves the ability of the multi-AUV system to deal with complex underwater environment.Finally,in order to solve the task assignment problem of heterogeneous multi-AUV system in the case of energy supply,a task assignment model of heterogeneous multi-AUV system considering the surplus energy is constructed.A multi-round bundled combined auction algorithm is proposed to solve the model.In the bidding package construction stage,each AUV solves the optimal bid package to meet its own needs and submits it to the auctioneer using the variable dimensional chromosome genetic algorithm.This limits the number of bids submitted by the AUV and reduces the complexity of the auctioneer to decide the winner.Furthermore,a winner decision algorithm is designed to preserve the task preference of the AUV to the greatest extent and ensure the correlation between tasks.This algorithm improves the overall utility of the AUV system and reduces the energy consumption.