Research On Key Indicators Of Urban Ground Transit Operation Monitoring Based On Passenger Travel Choice

The proposed "transit priority" policy eases traffic congestion,air pollution and other issues.Carrying out monitoring of urban ground bus operations can discover bus operation problems in a timely manner,which is conducive to enhancing the attractiveness of public transport and thus improving the rate of bus sharing.Urban bus operation monitoring is a complex system.The determination of key indicators is the basis of bus operation monitoring.On the one hand,it can be used for the government's supervision and evaluation of public bus operations,and on the other hand,it can pave the way for subsequent bus operation monitoring.At present,the bus industry can realize real-time data collection and storage,and has accumulated a large amount of resident travel data and bus operation data.At the same time,data mining and machine learning algorithms are suitable for large-scale data processing and analysis,which provides data and technical support key indicators for urban bus operation monitoring.This paper intends to use machine learning technology to determine the key indicators of urban ground bus operation monitoring to solve the problems of low efficiency and poor effect of existing urban ground bus operation monitoring.From the perspective of passenger demand,based on objective travel data and subjective survey data,the key factors affecting the travel choices of bus passengers are determined,and the key indicators of urban ground bus operation monitoring are determined.First of all,analyze the factors affecting the choice of bus passengers' travel,and form an initial set of indicators for bus operation monitoring.According to different data sources,the indicators are divided into objective quantitative indicators and subjective qualitative indicators,which are quantified by numerical calculation and cloud model respectively.Then,the cumulative prospect theory model is introduced,transforming the actual value of each indicator into a cumulative prospect value based on passengers' subjective perception to characterize the passengers' travel choice behavior more realistically.After that,the relationship between the indicators of passenger service level and the average passenger flow is studied based on feature selection engineering and machine learning.In the key index determination model,a sequence backward selection algorithm was used to search for a feature subset,and the prediction error of generalized regression neural network was used as the evaluation criterion to determine the key index of urban ground bus monitoring.Finally,the model application is developed for different travel time scenarios,and the key indicators of urban bus operation monitoring in different travel scenarios are calculated.After testing,the model errors are 0.0287 and 0.0339 respectively,indicating that the model established by the key indicators can solve the problem well.The results of the model show that the indicators that need to be closely watched during two different periods in weekdays includes indicators of speed and reliability.At peak times,it is necessary to pay attention to the non-linear coefficient of the line and bus fares.During the pingfeng period,the waiting time and comfort indicators should also be focused.the application of the cumulative prospect theory model can improve the reliability of key index research results without reducing the accuracy of the model,making the key index research results closer to reality.This research is dynamic and extensible,which adds new content to the research of urban ground bus operation monitoring,and can be more effectively applied in practice.The research results of the key indicators of urban ground public transport monitoring can not only be used as the basis for subsequent research in the field of public transport,but also can be used to guide public transport operations and government supervision.It is helpful to improve the sharing rate of public transportation and alleviate the problem of traffic congestion,and has certain social value.