Online Strategy For Task Allocation Problem Of Sharing Platform

With the upsurge of sharing economy,many new and successful sharing platforms such as Airbnb or Uber have been emerging in recent years.Under the sharing platform,not only the service time of resource providers is different,but also the information of users'future demand?such as arrival time,start time and duration?is highly uncertain.How to more effectively achieve the optimal matching of resource supply and demand is the key to the operation of the platform.Therefore,from the perspective of platform owners,this paper analyzes and studies the task allocation strategy under the sharing plat-form.Most of the existing researches on task allocation need to satisfy the hypothesis that the future demand sequence is subject to a certain random distribution or random process.However,due to the diversity of uncertain factors in reality,it is difficult to accurately describe the future demand distribution.In order to avoid the disadvantages of traditional static optimization methods,which rely heavily on the assumption of distribution condi-tions,this paper uses the online theory and competitive analysis to study the task allocation problem of the sharing platform and obtains some research results.First of all,we discuss the research background of this paper in detail and lead to the main content.Then we summarize the task allocation of the sharing platform and the online theory.In addition,we sort out the operation mode of the platform and the the-oretical basis of the competitive analysis.In particular,the existing research on online task allocation often assumes that the revenue is directly proportional to the demand du-ration,ignoring the important fixed revenue part of the platform's revenue.Therefore,the paper focuses on the online theory and competitive strategy to expand the revenue structure of the previous research platform and increase the fixed revenue part.And then,we design the competitive task allocation strategy,so as to optimize the total profit of the sharing platform.Thus,it can provide theoretical basis for platform decision-makers and help them to make better competition strategy in the worst case.Subsequently,the lower bound of structural competition ratio of task allocation strategy is studied,and we designed two kinds of task allocation strategies.Finally,the structure and content of the paper are summarized and described,and the future research direction is proposed.Specifically,the work and conclusions of this paper are as follows:1.Research on the lower bound of structural competition ratio of task allocation strat-egy.The offline problem of task allocation in sharing platform is part of the task allocation problem.Therefore,in the derivation process of the lower bound of structural competition ratio,we use Yao's principle to build and derive the model by ingeniously constructing the demand sequence and probability distribution,and do not use any complexity assumption to deduce.It is proved that any algorithm of task allocation problem takes_1+f^ln?+2as the lower bound,in which?and f are the relevant parameters of the problem given timing.2.Research on task allocation strategy of sharing platform pricing with fixed income.This part starts from the online theory,combines the structural characteristics of platform task allocation,and establishes the online decision mathematical model of task allocation under the condition of uncertain demand under the premise that the platform pricing rules are constant for each fixed price and unit variable price.We design the algorithm and allocation scheme of the problem by deeply simulating the characteristics of the problem that users face when sharing resources through the platform,and prove that the strategy is in and When n and?taking different values,it shows different competitive performance ratio.Through analysis,the competitive ratio of the designed busy strategy is only related to parameters f and?,but not to other parameters.The strategy has a better competitive performance ratio.Finally,the conclusion is verified by numerical simulation experiment.3.Research on the deterministic algorithm of pricing and service duration of sharing platform.This is a deterministic algorithm based on the previous chapter.In the algorithm,the defects of the previous chapter are overcome by rejecting small tasks,and the threshold parameter??i?is introduced to increase the acceptability of some tasks with longer dura-tion,so as to optimize the previous chapter.Through the analysis of competition ratio and numerical examples,it is concluded that the strategy has better competition performance.In addition,in the actual operation process of the platform,the decision-makers can adjust the threshold reasonably according to the actual situation to obtain more benefits.In this paper,the online theory is applied to the operation and management of plat-form,which is a useful extension of the online competition theory.In addition,the conclu-sion of this paper is helpful to the decision-making of task allocation of the shared service platform,and has important reference significance for the operation of the platform.