Study On Application Of The Highway Travel Time Prediction And Traffic Status Identification Based On Toll Data And Data Mining

As an important part of the highway,the highway toll system stored in the charging data more and more,with the continuous improvement of China’s highway network.At the same time,ITS is developing rapidly because the artificial intelligence and computer technology are applied to the field of transportation.How to dig out the valuable information from the massive fee data is an important research content of ITS.This paper uses the algorithm in machine learning to analyze the toll data,and carries on the application research of the expressway travel time prediction and traffic status judgment.This research provides the basis for the operation and management of highway,but also for the traveler to find out the highway traffic conditions to provide data support.Data processing is the first step of data mining.The data quality can affect the effect of data mining.First the data needs to be cleaned up,integrated and extracted to remove the obvious error data,the rest is valid data.Then,according to the theoretical knowledge of traffic flow,calculating the traffic flow parameters statistics that include travel time and OD flow.Those traffic flow parameters will be used in the next.The travel time is an important reference for expressway traffic.This paper studies the application of wavelet neural network in the prediction of travel time based on the charging data.Firstly,modifying the original travel time according to the characteristics of the expressway.Then,selecting the appropriate variables such as travel time of the previous period as input to the wavelet neural network.Lastly,adjusting the model parameters and getting the suitable parameters to obtain the optimal prediction results.The prediction model based on wavelet neural network is more ideal than BP neural network.In this paper,a discriminant model based on fee data and improved fuzzy clustering algorithm is proposed for the determination of expressway state.Combining the fuzzy clustering algorithm with the traffic flow theory,the selection of the initial clustering center is selected according to the highway traffic flow theory so that the amount of calculation is reduced and the efficiency is improved.The traffic condition index is calculated by weight averaging method for membership degree matrix obtained by fuzzy clustering model,the traffic condition index can reflect the highway traffic status.Judging the traffic status of the expressway sections and analyzing the result.For each section,this paper has corresponding case-based analysis of toll data.Through the example analysis to verify the precious,and for the practical application of research results provide a reference.