Research On Bus Network Optimization For Uncertain Demand

Conventional public transport has the characteristics of small investment,quick effect,easy adjustment of lines,large capacity of single vehicles,low per capita energy consumption and per capita consumption of pollutants,etc.Therefore,the development of public transport is one of the most effective ways to solve urban transport problems.The optimization design of the bus line network is based on the data of bus demand,which is constantly changing.Accordingly,the improvement of the line network's adaptive ability of changes in bus demand is a necessary condition to improve service level and reliability of bus services.Currently,the bus line network design is mostly based on fixed bus demand,not considering the bus demand changes due to time change,service dependence,randomness,and other factors,which will to some extent affect result validity of the optimal design of the bus line network.This article first introduces some basic theories of bus transit network optimization,analyzes the uncertainty of bus demand forecasting,the level of bus service and the reliability of bus services,and the impact of bus demand elasticity function on bus demand.Assuming that the transit demand obeys a certain normal distribution,the Monte Carlo method is used to simulate the transit demand matrix randomly.From the perspective of passengers,public transport companies,and society,the passenger travel time,the vacancy time of public transport vehicles,the time difference between bus travel and private car travel time,and bus fuel consumption are four objective functions.The three-party benefit weighting and maximum line network optimization objectives are used.A multi-objective optimization model of public transport network with uncertain demand was constructed to improve the adaptability of public transport network to uncertain public transport demand.Entropy weight coefficient method is applied to each objective function to give a certain weight ratio,making it an objective function to solve,to reduce the difficulty of the model solution,according to the different dimensions of each target using the economic transformation coefficient for the unified quantification of each target,Different target values are unified as the manifestation of the cost,and the solution ideas are designed and the optimal bus line combination is searched through the particle swarm algorithm.The effectiveness and rationality of the model and algorithm are verified through the Sioux-falls network test network,and the optimal bus line combination of the two scenarios of bus demand and single fixed bus demand is uncertain.Through comparing and analyzing the total target cost of the optimal bus line combination and the total passenger travel time cost of the two demand scenarios,the advantages and disadvantages of the optimal bus line combination under the two demand scenarios are discriminated.The results of calculation examples show that the indexes of bus line combination with uncertain demand are superior to the single fixed-demand bus line combination,and the designed bus line network can better adapt to the uncertainty of demand.At the same bus supply level,the bus line network with uncertain demand can provide higher service levels,and the bus line network is more efficient in responding to changes in bus demand.