Research On Auction-based Allocation Mechanism For Emergency Resource

Large-scale emergencies have occurred frequently in recent years,resulting in huge casualties and serious economic losses.The fundamental purpose of emergency management is to minimize the casualties and socio-economic losses caused by emergencies.In emergency management,emergency resources are the material basis,and their allocation directly affects the timeliness and effectiveness of emergency management,so it is of great significance to carry out research on the allocation of emergency resources.The problem of emergency resource allocation has obvious complexity,timeliness and dynamics.The existing model used for emergency resource allocation is that there is a central node to make decisions based on prior knowledge,and the emergency resource allocation scheme is determined in advance through planning.This model cannot deal with emergencies in a timely and dynamic manner,lacks adjustment methods and participants passively accept the results of allocation,with low participation enthusiasm.The auction is an effective way to allocate resources,which has the characteristics of high motivation,strong real-time performance and good robustness.Through the information interaction between bidders,the simulated item bidding process allows participants to select and bid on resources,achieving effective and fair resource allocation.Based on this,this paper studies the problem of emergency resource allocation based on auctions.Problem modeling,auction mechanism design and algorithm design are carried out for the two resource allocation scenarios of whether the demand for emergency resources is determined.The main research contents are as follows:(1)Problem modeling and algorithm design of the resource allocation problem based on the combinatorial auction under the determination of emergency resource demand.Considering the heterogeneity of emergency resource types and the differences of the participating allocation objects,the combination auction is applied to the emergency resource allocation to solve this problem.A model of emergency resource allocation based on auctions is established in this paper,in which different kinds of emergency resources are modeled as different auction items,demanders with emergency resource needs are modeled as completely rational auctioneers,and suppliers with idle emergency resources arepa modeled as completely rational bidders.The auctioneer’s goal is to maximize the service revenue from completing the emergency management task under budgetary constraints an resource demand constraints.The bidder’s goal is to maximize their utility by declaring the best combinatorial biddings of emergency resources to participate in the auction under the condition of satisfying individual rationality.A VCG-based auction mechanism is designed to maximize the utility of auctioneers.Theoretical analysis shows that the proposed mechanism satisfies the incentive compatibility of the dominant strategy and individual rationality.The incentive compatibility of the dominant strategy makes it the optimal action strategy for the completely rational bidders to pursue their utility maximization by adopting the real bidding strategy,so as to ensure that bidders will not conceal the price and increase the revenue of the auctioneer.Individual rationality makes the utility obtained by each truthful bidder no less than the utility he would have obtained if he did not participate in the auction of the emergency resource allocation,thus ensuring that bidders have the motivation to join the auction.Solving the winner set problem of the auction mechanism is proved to be an NP-hard problem.Intelligent optimization search algorithms are designed to implement the mechanism in this paper.(2)Problem modeling and algorithm design of the resource allocation problem based on the combinatorial auction under the uncertainty of emergency resource demand.Different kinds of emergency resources are modeled as different auction items,demanders with emergency resource needs are modeled as completely rational auctioneers,and suppliers with idle emergency resources arepa modeled as bounded rational bidders.Considering load balancing constraints of the resource allocation,a model based on stochastic optimization and a model based on robust optimization are proposed on the basis of the allocation model of emergency resource demand determination to solve this problem.The stochastic optimization model assumes that the probability distribution of the quantity variable of the demand for emergency resources is known,and the goal is to maximize the expected net income of the demander of emergency resources.The robust optimization model assumes that the uncertain quantity of emergency resource demand can be expressed by affine of some major uncertainties.The box uncertainty set is used in the robust optimization model.The goal is to maximize the net benefit of the demander of emergency resources in the uncertain set.A first price sealed auction mechanism is designed as the auction mechanism of the problem model,and the proposed mechanism satisfies individual rationality.Solving the winner set problem of the auction mechanism is proved to be an NP-hard problem,and intelligent optimization search algorithms are designed to implement the auction mechanism.This paper introduces the auction theory into the problem modeling of the emergency resource allocation,establishes auction-based resource allocation models for the scenarios of whether the demand for emergency resources is determined,and designs algorithms to implement the auction mechanism.This work aims to solve the weaknesses in the current optimization decision-making process of the emergency resource allocation and providing new ideas for realizing scientific,efficient,fair and reasonable allocation of emergency resources.It provides theoretical and decision-making references for emergency management departments in emergency resource allocation.