Research On Optimization Of Urban Bus Network Based On The Perspective Of Resilience

In recent years,the increasing passenger flow of urban public transport has led to huge safety risks in bus operation.In particular,different bus networks have different responses to emergencies.Some bus networks are paralyzed after an emergency,unable to quickly digest the adverse effects.Others can overcome adverse shocks quickly.In essence,there are differences in the resiliency of urban bus network: those with strong resiliency have strong adaptability and recovery ability,while those with weak resiliency have relatively backward reaction ability and longer recovery cycle.Therefore,it is imperative to strengthen the research on the resiliency of urban bus network.In order to improve the resilience of urban bus network,this paper uses intelligent optimization algorithm and traffic flow equalization algorithm to optimize the bus stop scheme.First,the paper analyzes the scientific definition of bus network resilience on the basis of literature research.Based on existing research,this paper defines the resilience of bus network as the weighted function of average travel time and average congestion.Then,the bus card swiping data,basic data of line stations and bus GPS data are preprocessed,such as data cleaning and index calculation,to provide data support for subsequent models and algorithms.In terms of public transport demand forecasting,using the method of deep learning instead of multiple regression,the growth rate of traditional method,such as depth of hybrid neural network OD passenger flow forecast model is established,and by using the super parameter optimization of particle swarm optimization,at the same time choose ARIMA,CNN,LSTM,GRU helped four models were analyzed,through the model performance is found,When the step size is long,the deep hybrid neural network model has high prediction accuracy and excellent stability.Secondly,based on the scenario in this paper,an algorithm of line OD passenger flow equilibrium is presented.Generalized costs such as waiting time and trip time are introduced,and the Logit travel choice model based on stochastic utility theory is used to evenly distribute OD passenger flow of line network,which verifies the convergence of the algorithm.Thirdly,the single-objective particle swarm optimization algorithm is designed with the goal of maximizing the toughness of the line network,and a case study of Beijing Special No.9 bus is carried out.The results show that the resilience of the optimized bus is improved by about 61.81% compared with that of the whole line and the whole station before optimization.Finally,with the introduction of multi-objective optimization idea,a multi-objective particle swarm optimization algorithm with the maximum line network toughness and the minimum generalized travel cost was designed to optimize bus routes to improve the toughness.The toughness of the Pareto optimal solution increased by 55%on average and the generalized travel cost decreased by 71% on average compared with that before optimization.Finally,the paper analyzes the influence of the toughness coefficient of the network,the operating cost of the bus company,the social interaction and the flow of stops respectively.The results show that :(1)the coefficient will significantly affect the toughness of the network,and the greater the C1 is,the greater the proportion of improvement of the toughness is.(2)The higher the station stop rate is,the larger the average travel time of the network is,and the higher the generalized cost is.(3)The introduction of cars will improve the optimization effect of crowding more obviously,indicating that the improvement of toughness may be more significant in practical application.(4)The larger the traffic is,the more times the station stops,and the smaller the traffic is,the more times the station stops,indicating that the line network optimization scheme has a certain interpretability.The research results of this paper can enrich the prediction model system of urban bus passenger flow,provide a basis for analyzing the impact evaluation of the new line on the existing road network,and make the route formulation and adjustment more reasonable.The optimization of the bus network by the intelligent algorithm can make the optimization of the bus network more scientific,rigorous and effective,enhance the toughness of the bus network,and improve the disaster tolerance,resistance and recovery ability of the large-scale urban bus network in the face of emergencies.