Research On Site Imbalance Adjustment And Flow Forecasting Methods Of Shared Bicycle System

The emergence of shared bicycles provides a new way of green travel,which provides great convenience for people's lives.At the same time,some problems are inevitable,the most obvious of which is the rental and return imbalance of shared bicycles between the same site.Therefore,in order to solve this problem,it is necessary to constantly adjust the balance state of the site,so that any site can rent and return the shared bicycle at any time,improve the utilization of the shared bicycle,save the user to rent and return the shared bicycle time.In this paper,we mainly study the unbalance adjustment of the shared bicycle system site and the prediction in the shared bicycle system.The research of this paper is summarized as follows:1.Aiming at the problem of balance of supply and demand of bicycles in shared bicycle stations,this paper proposes a method for adjusting the imbalance of shared bicycle stations based on Pareto multi-target selection.Firstly,the collected shared bicycle usage data is analyzed,and the quadruple method is used according to the frequency of shared bicycles.All shared bicycle stations in the city are divided.Secondly,in the dense area where the bicycle is used,the station in the area is unbalanced and the imbalance is statistically calculated.Finally,for the sites with serious imbalance between supply and demand,centered on it,the imbalance between the neighboring sites within a certain range is determined,and the reverse unbalanced sites are selected.The Pareto multi-target selection method is used to select from these sites.The candidate site for its imbalance adjustment,two-way imbalance adjustment,to solve the problem of site imbalance,and verify the effectiveness of the algorithm through comparative experiments.Experimental results show that the proposed adjustment algorithm is superior to other algorithms.2.To solve the problem of unbalanced stations,a better way is to predict the distribution of shared bicycles in advance.Therefore,this paper proposes a hierarchical prediction method for shared bicycle systems based on two-level spectral clustering to predict shared bicycle traffic.Firstly,the two-level spectral clustering algorithm is used to cluster the shared bicycle stations into multiple classes,which takes into account the geographical location information of the shared bicycle stations and the migration trend between the shared bicycle stations.Clustering is performed according to the geographical location information of the site at a low level,and repeated clustering is performed at a higher level according to the migration trend and the clustering result of the geographic information of the low-level site until convergence.Secondly,the gradient-enhanced regression tree(GBRT)algorithm is used to learn and predict the total number of shared bicycle rentals,and the multi-similarity inference model(MSI)inference model is used to study the rental ratio,migration trend,and interval learning and prediction.This allows you to derive the rental and return of shared bikes at each site.In this paper,the accuracy of the prediction model is verified by experiments,and compared with the current widely used prediction models.The experimental results prove that the prediction model proposed in this paper has obvious advantages.