Research On Urban Travel Behavior And Application Based On Social Computing

With the accelerated process of global urbanization,urban travel behaviors are becoming increasingly complex,accompanied by pressing challenges such as urban traffic congestion,air pollution,and energy waste.Studying urban travel behaviors and patterns is of great significance for optimizing urban transportation systems and enhancing urban sustainability.Traditional research on urban travel mainly relies on methods from transportation engineering and urban planning,focusing on the construction of transportation facilities and the optimization of traffic flow,while neglecting the complexity and dynamics of individual behaviors and failing to understand the relationship between travel behavior and urban structure from a macro perspective.Meanwhile,the increase in urban population and urban expansion have led to the diversification of travel patterns and the complexity of urban spatial structure,posing new challenges to the study of urban travel behaviors.Social computing,by combining computer science and social science,provides effective methods for researching and analyzing the increasingly complex urban travel behaviors.This paper mainly accomplishes the following three aspects by mining human urban travel datasets.Firstly,this paper studies the spatiotemporal characteristics and access patterns of taxi operations,and models them.On the one hand,previous studies have overlooked the exploration of taxi off-load processes and have not attempted to uniformly model the different operating states of taxis.This paper extracts time and distance features of on-load and off-load processes from taxi driving data and finds that they exhibit different statistical distributions.From a profit-driven perspective,based on the Langevin equation,this paper proposes two mathematically unified models,replicating the spatiotemporal characteristics of taxi on-load and off-load processes.On the other hand,this paper focuses on the access processes and mechanisms of taxis to different locations.It is found that,at the collective level,taxis show significant heterogeneity in their access frequencies to different areas.At the individual level,taxis exhibit consistent patterns of access to different regions.Compared to human individual mobility,taxis have a stronger exploration tendency and a weaker preference return tendency.By establishing a taxi exploration-preference return model,this paper successfully simulates the access patterns of taxis to different areas and reveals the critical role of the ranking distribution in the access process.These findings deepen our understanding of taxi work strategies and provide scientific guidance for improving and optimizing urban residents’ travel by taxi.Next,this paper investigates the characteristics of urban mobility networks and the influence of urban structure on urban travel behavior.The urban mobility network is constructed by treating urban area locations as nodes and human mobility trips as edges,encapsulating spatial attributes.It is found in this study that both the nodes and edges of the urban mobility network exhibit varying degrees of heterogeneity.Interestingly,the node degree distribution of the mobility network presents a unique two-section power-law distribution,with the two segments corresponding to the central and non-central areas of the city,indicating a distinct hierarchical phenomenon in urban travel behavior between two areas.Additionally,a novel approach to predicting urban travel flow is proposed in this paper.Building upon the traditional gravity model,a grid-based traffic flow prediction model is introduced,capable of accurately predicting travel flow between various regions of the city.Through community detection on the mobility network,this study also reveals the close relationship between urban mobility communities and administrative boundaries and road networks.Two hypotheses are proposed,namely,that the distribution of taxi passenger distances and the structure of urban fundamental communities influence community evolution,and a taxi community mobility model is established,reproducing the phase transition phenomenon in the evolution of the largest mobility community with trip distance.These findings can assist urban managers in achieving more effective control over urban travel flow and optimizing urban structural layout.Finally,this paper investigates the optimal and minimum fleet problem in urban transportation.Establishing an efficient,low-energy-consumption,and low-emission transportation system is one of the daunting challenges facing modern cities.Based on travel demand data,this paper proposes an optimal and minimum fleet algorithm aimed at further reducing idle travel time and pollution emissions of the minimum fleet.The optimal and minimum fleet algorithm determines the optimal fleet scheduling by establishing a vehicle scheduling network and leveraging classical network flow problems in graph theory.The algorithm is applied to travel datasets from three cities.Experimental results demonstrate significant reductions in connection time and pollution emissions,with the total connection time of the minimum fleet reduced by over 50% compared to the original algorithm.It is found in this study that both fleet size and connection time exhibit linear relationships with travel demand,increasing with the volume of travel demand.The city size also affects the slope of these linear relationships,with smaller cities having smaller slopes,making efficient fleet operation easier to achieve.These results underscore the enormous application potential of the optimal and minimum fleet algorithm,which can help cities establish demand-driven,non-profit,and eco-friendly transportation systems.