The Research On Dynamic Scheduling Of Shared Bicycle Based On Demand Forecast

For the past few years,the number of motor vehicles in China has grown rapidly,resulting in increasingly prominent urban traffic problems.As a green travel method in the sharing era,shared bicycles have suddenly emerged,which can not only ease the pressure of urban traffic,but also help solve the "last mile" problem.However,while the shared bicycle system is developing vigorously,there are situations of "difficult to borrow a car",which seriously restricts the continuous development of the shared bicycle system.Based on this,this article analyzes the characteristics of shared bicycle travel demand,and then further studies the prediction of shared bicycle demand and the optimization of scheduling paths.This article focuses on analyzing the characteristics of shared bicycle travel demand from four aspects: traveler characteristics,traffic characteristics,time distribution characteristics and spatial distribution characteristics.Furthermore,combined with the spatial and temporal distribution characteristics of shared bicycle travel demand,this article proposes a BP neural network algorithm-based shared bicycle borrowing demand forecasting method,which predicts the borrowing demand of the same attribute on the corresponding period of the forecast day.According to the predicted borrowing demand,the scheduling demand can be determined,and then the two-stage model of multi-objective dynamic demand vehicle scheduling based on user satisfaction is studied,with the lowest scheduling cost and the highest user satisfaction as the goals.Real-time optimization stage scheduling model.Among them,the maximization of user satisfaction translates into the minimization of the penalty cost for exceeding the scheduling time window,thereby transforming the multi-objective problem into a single-objective problem;the concept of virtual scheduling center is introduced in the real-time optimization stage,thereby transforming the dynamic problem It is a static problem.According to the characteristics of the model,a hybrid ant colony algorithm combining ant colony algorithm and genetic algorithm is designed to solve the model.This article uses the above research methods,to predict a representative drop point at various times of the day,and to solve the pre-optimized scheduling program and real-time optimized scheduling program of the drop-out point within the operating area.The results show that the demand forecast results basically match the actual demand situation,and the scheduling model is also solved to obtain good results,which verifies the rationality and advancedness of the shared bicycle dynamic scheduling method based on demand forecast researched in this paper.