Optimization of Average Travel Time of Passengers in Tehran Metro Using Data Mining Methods

In the design and development of public transportation systems such as urban railways, not only the route and location of stations, but also the timetable of fleet movements must be considered. Train timetables are an important factor as they influence customer satisfaction, metro operating costs, and environmental health; as a result, train timetables optimization increases the service quality. In this timetable optimization, train stopping time at the station as well as passenger waiting time would be taken into account. In time optimization research, mathematical analysis and simulation of data mining algorithms have been used so far for general changes in timetables. In this article, the data are examined in detail in order to find significant differences with other data. In this paper, the data of delayed trips in Tehran metro have been examined using data mining analysis methods. The Discriminant Analysis method has been used to identify delayed trips with significant differences after a relative understanding of important features of the dataset. Considering the power of the genetic algorithm to achieve an optimal solution, the approach proposed in this paper is to provide a solution to combine this algorithm and discriminant analysis method. The result of this paper is 40 final chromosomes, indicating trips that exhibit a significant difference in latency, compared to other trips. And according to the characteristics of these trips, the optimization can be done by changes in the timetable.