Research On Tide Positions’ Dispatching Strategy Of Shared Bicycles In Urban Morning Rush Time

With the explosive growth of the urban population and the number of automotive vehicles in our country,the problem of urban congestion has become increasingly serious.In addition to the choice of private cars,people mostly choose public transports such as buses and subways when traveling in cities,and the emergence of shared bicycles has changed our lives.It alleviates traffic congestion and satisfies the short-distance demand of the "last mile",but it still exposes some problems such as nowhere to stop,random parking,occupying blind lanes and no bicycles to ride.This article mainly starts from the two perspectives of users’ scheduling and merchants’ scheduling,and aims to solve the problem of no shared bicycles and piles of shared bicycles in the morning peak,and finally obtains a reasonable scheduling strategy and scheduling path.The main research work is as follows:1)Based on the trip data provided by the 2021 Digital China Innovation Contest(DCIC2021),Xiamen City from December 21,2020 to December 25,2020 from 6:00 to 10:00.Firstly using geohash encoding to convert the latitude and longitude data into a string.Then identify the faulty bicycles by the length of the lock status change time and call attention to the merchant.Secondly this article researches users’ preferences and time-space rules for using shared bicycles to travel.Meanwhile using data visualization and data mining.The analysis shows that the demand and saturation of shared bicycles have obviously temporal and spatial differences,which means that the demand of shared bicycles will peak at around 8:20,and is mostly in workplaces,subway stations,commercial areas,etc.At the meantime,bicycles are mainly to meet the needs of people for short-distance travel.2)From the users’ perspective,identify top 30 areas with the tidal phenomena in the morning on weekdays.And then use the KNN model with constraints to design the optimization plan of tidal positions,which means that the new recommended parking spot is not a tidal one and the new one is on the side of the road with the original one.And actively guide the user to the nearest parking spot.Cut peaks and fill valleys to alleviate congestion problems at tidal positions.Among them,the HNSW algorithm is used to reduce the complexity of the issue.3)From the merchants’ perspective,perform k-means clustering on the top 30 positions for regional division with the highest demand and saturation,transform the global scheduling into a local scheduling problem,which can reduce the complexity of the issue.Secondly establish a model with the lowest scheduling cost as the objective function.Then using tabu search algorithm,simulated annealing algorithm,and ant colony algorithm to solve the problem.And finally get a more logical scheduling strategy obtained by ant colony algorithm,and the distance is calculated by the haversine formula.