Research On Demand Calculation And Dynamic Scheduling Of Shared Bikes Based On Spatio-Temporal Changes

Shared bicycle is a new type of flexible,convenient,economical and green transportation,which has become an important choice for travelers because of its small size,large number and easy operation.This thesis investigates the spatial and temporal demand of shared bicycles,and explores the mechanism and dynamic control scheduling scheme to improve the utilization and turnover rate of shared bicycles.The spatial grid is set up with the function of "fishing net function" of Arc GIS,and the fusion of shared bicycle rental,return and P0 I data is realized in the spatial grid;the heterogeneity of shared bicycle demand is studied in time and space dimensions,and it is found that the daily borrowing and returning quantity of shared bicycles is basically stable,and the spatial demand distribution does not fluctuate much;however,the borrowing and returning quantity and spatial distribution of 24-hour However,the heterogeneity of the 24-hour borrowing and returning volume and temporal and spatial distribution is stronger,mainly related to the number and diversity of POIs,and the above factors are taken as the main factors for the analysis of the influence mechanism of shared bicycle demand.Using shared bicycle weekend data,the time series algorithm and DBSCAN spatial clustering model were fused to cluster the time series of a single day in the time dimension.The time was divided into 6 sequence segments to explain the temporal heterogeneity of shared bicycle usage patterns on weekends;The spatiotemporal cube is constructed to conduct spatiotemporal thermal analysis on the grid cell,and the shared bicycle use space is divided into 12 grid aggregation regions according to the principle of homogeneous regions and the nearest neighbor selection model to explain the spatial heterogeneity of the shared bicycle use law;Use the above achievements as the basic time units and regions for the research on dynamic scheduling of shared bicycles.Based on the clustering results,set up a control group and establish time weighted regression models(TWR),geographic weighted regression models(GWR),and spatiotemporal geographic weighted regression models(GTWR)from both temporal and spatial dimensions,respectively,to analyze the spatiotemporal relationship between shared bicycle usage demand and POI distribution;Finally,the rationality of the clustering results was verified,and it was found that the accuracy and stability of the GTWR model were better than those of the TWR model and GWR model;The combination of clustering results and spatiotemporal weighted models provides a more detailed explanation of the impact mechanism of shared bicycle demand in the spatiotemporal dimension.Finally,the shared bicycle demand is calculated based on the impact coefficient.Based on the time series period,grid aggregation area,and calculation results obtained in the previous text,a dynamic allocation strategy for shared bicycles between grid aggregation areas is proposed based on the classic production and sales imbalance transportation problem theory in operations research.Combined with the actual shared bicycle scheduling process,a dynamic allocation plan for shared bicycles is established;The obtained allocation plan can effectively solve the problem of spatiotemporal heterogeneity of shared bicycles,achieve real-time adjustment of shared bicycle scheduling operations,and improve the utilization and turnover rate of shared bicycles.Select Futian District,Shenzhen City as an example area to develop an actual allocation plan for shared bicycles within that area.