Public Transit Network Design Based On Travel Demadn

With the continuous expansion of urban size and population growth,the vigorous development of public transport has become the inevitable way to ensure the transit service of urban residents.Currently,metro lines have been built in many larger cities,but the public transit demand in many cities still depends on bus transit systems.Bus travel still plays a dominant role in many cities in our country.In the city with subway systems,the normal bus transit also bears a lot of travel demands from the subway stations.The design of public transportation network should be based on the distribution of passenger flow,preferably the starting and ending points of all the routes of urban public transportation,the optimal route planning and the interconnection between different routes.The main task of Urban Transit Network Design is to design public transit network structure and departure frequency based on existing urban road network and public transportation capacity,which can reduce the number of passenger transfers,reduce the average waiting time and improve the public transportation service level.In this paper,a bi-objective programming model is established with the objective of passenger travel cost and bus operating cost,and solved by using non-dominated sorting genetic algorithm.By doing this,this paper tackle the improper selection of weight in the literature,in which the Urban Transit Network Design are solved with a combined objective with weighted multiple objectives.Based on the distribution of transit demand,a novel initial transit network generation algorithm is proposed to reduce the computation time and computation resources required to search for a better transit network structure and departure frequency.The formulation of a transit route set alternating heuristic has expanded the search space with fixed number of routes to the search space with variable number of routes.Finally,the model and algorithm in this paper are applied to Mandl’s Swiss road network.The result obtained by this paper is compared with the latest research results in the literature.The results show that the proposed method can optimize the structure of the transit network and reduce the total travel time and the size of the fleet while meeting all the travel demand.The result has improved the utilization of public transportation resources,which is more conducive to the planning of public transportation network.The analysis also shows that the proposed algorithm can provide approximate Pareto solutions for decision-makers to choose the best solution,which is more in line with the actual situation of bus network planning than the traditional single-objective model algorithm.