Research On The Optimization Of O2O Takeaway Crowdsourcing Distribution Path Considering Rider Familiarity

In today’s era of rapid development of the Internet,O2O（online-to-offline）is favored by the Chinese people due to its outstanding convenience,which also directly promotes the development of instant delivery.Various types of O2 O platforms are constantly appearing in front of people.Among them,food delivery,as a strong related industry of instant delivery,plays an irreplaceable role.In today’s increasingly accelerated pace of life,more people prefer to place orders on the platform to solve their own meals in order to save working time.question.Riders,as the key link of "Internet + service industry" and "intelligence + logistics",are a flexible professional group formed with the vigorous development of platform economy and digital economy.In recent years,with the increasingly widespread application of online food delivery services in consumers’ daily life,riders connect users and stores through delivery,becoming the capillaries connecting cities and playing an increasingly important role in urban life.Riders connect users and merchants through distribution centers.As an important capillary channel connecting cities,they play an increasingly critical role in urban life.In this study,the rider’s familiarity with each location is known and can be quantified.The rider’s familiarity can affect the delivery speed,which in turn indirectly affects the delivery time.This paper also introduces the rider’s familiarity area and the concept of unfamiliar area,the rider’s familiarity with the familiar area is high,which positively affects the delivery speed and shortens the delivery time.The rider’s familiarity with the unfamiliar area is low,which has a negative effect on the delivery speed and increases the delivery time.In addition,in order to improve the overall benefits of riders,this study adds constraints that limit the maximum delivery volume of riders to the model to avoid situations where individual riders with strong abilities take more orders,while other riders are assigned to a small number of orders,so that each rider can be assigned to a similar number of orders to achieve a fair distribution of order quantity,delivery time and delivery revenue.And build the appropriate VRP mathematical model to solve in Gurobi solver.Design an algorithm with a high degree of fit for this research-TAVNS（two-stage adaptive variable neighborhood search algorithm）.On the basis of the basic variable neighborhood algorithm,a two-stage search and adaptive mechanism is added to speed up the solution time of the algorithm,and the unique neighborhood structure and search strategy of this study are designed to enhance the solution efficiency of the algorithm.Finally,the improved Solomon benchmark example is used to generate examples of different scales,the results of the solver are compared vertically,and the results of the TAVNS algorithm after removing different operators are compared horizontally.The results show that the TAVNS algorithm performs well in the calculation example,and the designed operators and search strategies have played their best role.