Designing a Driving Cycle in the City of Semnan with Data Collection Using Chasing Vehicles and Clustering with the K-means Algorithm

The driving and traffic cycle in cities is one of the most important and complex challenges in urban management. For the data collection process, the chasing vehicle moved in a certain route in the city through the selected time period, and data collection was conducted with software installed on a mobile phone. This research investigated the driving cycle in Semnan city using the chasing method and the K-means clustering technique. The distance covered during data collection was approximately 12 km for each route. Additionally, the total distance covered for the data collection in this study was approximately 130 km. The car model information, age, and gender of chased drivers were recorded as the influential parameters. Then, using the K-means technique, the data collected for different routes were analyzed to extract the behavioural patterns in the routes of Semnan city. Furthermore, in this study, the appropriate number of groups for data clustering was investigated, and considering the standard deviation of the data, it was concluded that the optimal number of clusters was 5. Increasing the number of clusters, despite improving the accuracy of calculations, led to longer processing times. Therefore, by clustering with 2, 3, 4, 5, and 6 groups, it was observed that these numbers of clusters improved by approximately 40, 80, 90, 98, and 99%, respectively, indicating that increasing the number of clusters showed a significant improvement in the results. These clusters included urban roads, secondary roads, highways, motorways with light traffic, and start-stop. This information can help city officials and traffic managers with traffic and safety improvement programs, according to the specified patterns, and provide a deeper understanding of driving behaviour in Semnan city.