Font Size: a A A

Reserch On Public Bysicle System Forcast And Allocation Problem

Posted on:2017-06-25Degree:MasterType:Thesis
Country:ChinaCandidate:C W HeFull Text:PDF
GTID:2322330485490970Subject:Engineering
Abstract/Summary:PDF Full Text Request
Because public bicycle has the characteristics of public use,convenience,easy to be managed,low cost,it has received great development at home and abroad.Only in the public bicycle system can provide enough vehicles and plenty of free parking spaces,its benefits can be fully played.To ensure this,the public bicycle system needs the help of the allocation of vehicles.Determining Efficient allocation not only need to optimize the deployment of path,we also need to forecast the number of borrowing and returning bicycles to ensure Site has been deployed before the "problems" show up.Therefore,it has important significance to predict the number of lending and returning public bicycles and to research the deployment of bicycle.First of all,this paper analyzes the domestic and foreign public bicycle allocation theory and demand forecast research status,to find the deficiency of the theory of the emergence of the study,determine the research direction and objectives.Secondly,according to the Suzhou Gaoxin zone 9 public bicycle station for 32 consecutive days,the time interval for the 10 min borrowed car data,analysis of public bike,car demand in time and space distribution characteristics.According to the public bicycle by the advantages,disadvantages and applicability of the also car demand data periodicity and randomness,combined with a variety of short-term traffic flow forecasting method of determined by time series method in the SARIMA model(seasonal difference divided into auto regressive moving average model)to model and forecast,the elaborate process modeling of stationary test,differential smoothing,model order and other steps,and to achieve a fitting and prediction using JMP software.Taking into account the different deployment cycle,the data of 20 min,30min and 40 min time interval are predicted by the same method.The prediction results show that with the increase of time interval,the prediction accuracy is getting worse and worse,and the SARIMA model can only accept the prediction of 10 min,20min and 30 min time interval data.Then,analyzing the problems existing in the public bicycle vehicle scheduling,and puts forward a vehicle scheduling method based on demand forecast.In this method,the allocation threshold and the optimal ratio of the pile,the allocation and the allocation area are analyzed and explained.Finally,the choice of allocation path into the traveling salesman problem,determine the specific route dispatching.
Keywords/Search Tags:Public bicycle, Short-term traffic flow prediction, SARIMA model, Dispatching method
PDF Full Text Request
Related items