Research On Urban Resident Trip Decision Model Considering Risk Preference In Profit And Loss

With the rapid development of economy and society,the rapid advancement of urbanization and motorization,the continuous improvement of infrastructure and comprehensive transportation systems,the continuous expansion of the city’s scale,the continuous increase of service functions undertaken,and the increase in population and passenger flow,The resulting urban traffic congestion,frequent traffic accidents,and increasingly serious traffic pollution have become common problems faced by cities at all levels throughout the country and the world,and have become bottlenecks that restrict the further development of society and the economy.The traditional method of resolving urban transportation problems solely by increasing investment and carrying out large-scale transportation infrastructure construction is gradually diminishing the marginal benefits of optimizing the comprehensive urban transportation system,and it is no longer suitable for the rapid development of urban transportation.The focus of urban transportation planning is changing from the original large-scale transportation infrastructure construction to the effective and accurate management of travel needs.Research on the decision-making process of urban residents’ travel is the key to grasping travel needs and formulating control measures.This paper selects the residents of the central urban area as the research group,and takes the decision of travel mode and departure time as the research object.By constructing the questionnaire of urban residents ’travel information and the questionnaire of residents’ travel preferences,the personal attributes,family attributes,travel Investigate the basic data of attributes and decision preference attributes,and obtain a total of 1,200 valid recycling questionnaires;use various open source geographic data platforms such as Gaode Geographic Open Platform and Open Street Map(OSM)to obtain various types of POI data and land use data based on transportation communities.The obtained original data is analyzed by correlation between variables,and the explanatory variables are reconstructed based on the entropy weight method / factor analysis method to achieve data dimensionality reduction.Through the analysis of the correlation between explanatory variables and structural variables,the upper and lower layers of the explanatory variables in the model construction are determined.Configuration.Based on the reconstruction and the definition of the explanatory variables of the configuration,in order to overcome the existence of the model construction based on the theory of random utility only,it is assumed that the traveler makes a rational decision and fully grasps the global information,which does not completely match the actual decision situation;the perceived value of the plan and The objective utility obtained by theexpectation theory is not completely consistent;lack of expressions of risk decision attitudes and decision preferences of travellers in uncertain environments;lack of rationalization of weights for extreme probability / utility schemes;and so on.This article uses the random utility theory to study travel behaviors.Based on the widely used NL model,the cumulative outlook theory is introduced to improve the model framework.By constructing reference points,profit and loss values,and value functions under profit and loss,the traditional utility function is improved,and the objective utility is subjective.The preferred schemes perceive value,and predict the scheme from top to bottom through the foreground value-cumulative foreground value.After the model construction is completed,the statistical indexes of the model and the prediction hit rate of the modeling group / testing group are tested;according to the calibration of the model parameters,the sensitivity of each explanatory variable-structure variable is analyzed based on the elasticity theory,and combined with descriptiveness The results of statistical analysis put forward measures to optimize the travel structure of urban residents.The research results show that the combined hit rates of the model group / test group are84.38% and 81.75%,which are significantly higher than the predicted hit rates based on the NL model alone(72.38% and 70.50%).The prediction of travel plans has better stability.The results of the sensitivity analysis showed that: monthly income,age,number of families,number of non-motorized vehicles in the family,number of motorized vehicles in the family,travel time,time value concept,monetary decision-making weight,satisfaction of private car /taxi trips,and service functions of the place of departure Factors such as transportation service functions have a significant impact on the choice of travel mode;education level,distance from bus stops,travel time,travel experience motorization requirements,transfer motorization requirements,bus travel satisfaction,full-day travel costs / time budget,Factors such as the residence service function of the departure place have a significant impact on the choice of departure time.