Research On Optimization Of Takeaway Distribution Route

In recent years,the scale of the takeaway market has continued to grow.User demands have expanded from the early simple food to the whole category,and the addition of fresh products has also added new vitality to the takeaway market.The transaction value of China’s takeaway industry has exceeded 600 billion in 2019,and the scale of takeaway platform users has stabilized,but the takeaway penetration rate is less than 15%,and there is still room for continued growth in the future.With the acceleration of the pace of life,consumers are increasingly willing to pay for convenience,and takeaway has become a common catering consumption habit.The current takeaway market has matured,the takeaway intelligent dispatching system has been gradually optimized,and the industry’s overall distribution efficiency has improved,but hidden dangers to traffic safety are common.On this basis,under the premise of ensuring the security of distribution,how to continuously improve the industry’s distribution efficiency and user’s takeaway experience,while reducing the distribution costs of the enterprise has become an urgent problem.At present,most of the domestic and foreign research on takeaway distribution is focused on the macro level,and there is not much research on distribution route planning of takeaway vehicles in actual scenarios.In the model construction,this thesis takes the takeaway vehicle routing problem of a single distribution center as the objection,uses the distribution costs and customer satisfaction as the dual goals,and establishes a penalty function to convert the customer time windows and the acceptable latest delivery time into penalty costs for calculation.In addition to the conventional constraints such as time windows and vehicle loads,two constraints that were rarely considered in previous research are added.One is the time-varying road network constraint.When the vehicle is traveling in the actual road network environment,the speed is dynamically changed.One is the road safety constraint.Each route has its own safety.The vehicle will preferentially choose the route with higher safety for distribution.The consideration of this constraint is also the innovation point of this thesis.Among them,the time-varying problem introduces a cross-period method that meets the first-in-first-out criterion to calculate travel time.Quantification of road safety is a key task.Introducing the consideration of influencing factors and evaluation methods for road safety in the field of transportation,and combining the characteristics of takeaway distribution to determine the danger value of the section under study by questionnaires,field trips and principal component analysis.Based on the above considerations,the multi-constrained takeaway vehicle distribution models with the lowest costs in non-time-varying and time-varying scenarios were established.According to the characteristics of the model,this thesis chooses the highly adaptive ant colony algorithm to solve,and optimizes the basic ant colony algorithm in three points.The first is to construct heuristic functions that consider both the distance and security between nodes.The second is to introduce a random variable into the transition probability selection rule to avoid the algorithm from stagnation.Third,in the processing of pheromone,referring to the maximum-minimum ant system,set the maximum and minimum boundaries of pheromone on the path,and only update the pheromone increment released by the ant who found the optimal path.After each loop,the 2-opt algorithm is incorporated to locally optimize the planned path.Considering the research scope of the thesis,taking the merchant self-delivery mode as an example,the takeaway distribution in the actual scenario under non-time-varying and time-varying situations were simulated,and the results were compared and analyzed.Then,taking the takeaway distribution under time-varying conditions as an example to compare the algorithms before and after the improvement,verified the necessity of the constraints considered in the model establishment and the effectiveness of the improved algorithm.At present,there are no effective policies and regulations to restrict takeaway distribution behaviors.The research on the optimization of takeaway distribution routes considering road safety can guarantee distribution costs and maximize customer satisfaction.At the same time,it can reduce hidden safety hazards and alleviate traffic safety.In the future,we will continue to deepen research and expand its application fields.