With the rapid development of the Internet and intelligent control technology,
intelligent transportation has become a research hotspot in building a smart
city. Under the background of intelligent transportation, it is particularly
important to effectively evaluate the rail transit as the framework of urban
public transport in this study, and fuzzy mechanism is introduced to optimize
the support vector machine (SVM), and on this basis, analytic hierarchy process
(AHP) and SVM are combined to improve the classification accuracy and improve
the rail transit operation safety evaluation index system. The experimental
results show that the classification accuracy of the fuzzy SVM combined with AHP
is above 85% on all the datasets, and it can effectively eliminate the
less-relevant indicators. In the actual evaluation of Shanghai Rail Transit
safety, the prediction accuracy exceeded 80% and the highest reached 94.51%.
Among them, the accuracy of management level and infrastructure were increased
by 24.1% and 18.34%, respectively, indicating that this method can effectively
screen the evaluation indicators. In the evaluation of Beijing Rail Transit, the
accuracy rate of the combined algorithm reaches 95.67%, with high classification
accuracy, which provides a reference direction for the establishment of the rail
transit operation safety evaluation system.
