Research On Dynamic Allocation Optimization Of Bike-Sharing Based On Data Analysis

Bicycle,which has been gradually crowded out by cars in the past two decades,is gradually returning to public view with the emergence of "bike-sharing craze".Bike-sharing is a product of scientific and technological innovation,and its emergence facilitates the effective solution of the "last mile" problem.In the early stage,operators made crazy efforts to seize the market by putting bike-sharing,which wasted a lot of resources and occupied a large number of public places.At the same time,there was a serious imbalance between supply and demand of bike-sharing in parking spots.Therefore,it is necessary to optimize the allocation scheme of bike-sharing to increase the turnover rate of each bike,meet users' travel needs with an appropriate number of bike-sharing,and avoid the phenomenon of large-scale investment in bike-sharing to continue.In this way,the situation of high input cost,low utilization rate of bike-sharing and imbalance between supply and demand of bikes can be effectively improved.Based on this,the research work of this paper includes the following aspects:(1)Based on the travel data of mobike in Beijing,the travel characteristics of users of bike-sharing are analyzed.In addition,by means of questionnaire survey,the user usage and satisfaction in the current bike-sharing market are studied,and the user satisfaction index of riding is obtained by statistical calculation.These results provide theoretical support for the subsequent travel prediction of bike users.(2)K-means clustering method was used to divide the delivery area of bike-sharing.On this basis,the travel demands of users with bike-sharing in each delivery area were dynamically predicted by using Xgboost algorithm.The expected travel demands of bike-sharing under ideal conditions were calculated by combining with the users' riding satisfaction index.In addition,by acquiring the longitude and latitude data of the starting and ending points of mobike,the user's travel distribution between the launching areas was obtained,and the expected travel distribution under ideal conditions was predicted.(3)The dynamic allocation optimization model of bike-sharing is established.From the perspective of operators,the demand of users and the cost of operators are considered,and the allocation amount of bicycles between different delivery areas in each period is determined to maximize the net profit of operators and satisfy users'travel needs.Finally,taking shijingshan district of Beijing as an example,the model is applied to obtain the relevant dynamic deployment scheme,which provides the relevant theoretical support for the deployment activities of bike-sharing operators.