Clustering Analysis of Multilayer Complex Network of Nanjing Metro Based on Traffic Line and Passenger Flow Big Data

Abstract

Complex networks in reality are not just single-layer networks. The connection of nodes in an urban metro network includes two kinds of connections: line and passenger flow. In fact, it is a multilayer network. The line network constructed by the Space L model based on a complex network reflects the geographical proximity of stations, which is an undirected and weightless network. The passenger flow network constructed with smart card big data reflects the passenger flow relationship between stations, which is a directed weighted network. The construction of a line-flow multilayer network can reflect the actual situation of metro traffic passenger flow, and the node clustering coefficient can measure the passenger flow clustering effect of the station on adjacent stations. Combined with the situation of subway lines in Nanjing and card-swiping big data, this research constructs the line network with the Space L model and the passenger flow network with smart card big data, and uses these two networks to construct the multilayer network of line flow. This research improves the calculation method of the clustering coefficient of weighted networks, proposes the concept of node group, distinguishes the inflow and outflow, and successively calculates the clustering coefficient of nodes and the whole network in the multilayer network. The degree of passenger flow activity in the network thermal diagram is used to represent the passenger flow activity of the line-flow network. This method can be used to evaluate the clustering effect of metro stations and identify the business districts in the metro network, so as to improve the level of intelligent transportation management and provide a theoretical basis for transportation construction and business planning.

Multilayer Seasonal Autoregressive Integrated Moving Average Models for Complex Network Traffic Analysis

“…Network traffic constantly changes, and patterns can emerge and disappear within seconds. Therefore, it is crucial to have real-time monitoring and analysis systems in place to detect and respond promptly to any anomalies or malicious activities [15]. Furthermore, network traffic is only sometimes generated by legitimate users and devices.…”

Improved Long-Term Forecasting of Passenger Flow at Rail Transit Stations Based on an Artificial Neural Network

“…Figure 3 also shows that the closer to the city center, the greater the value. It shows that the rail transit stations in the center of Beijing bear a large amount of passenger flow, which is of great signifi- (2) Cluster centrality (CC) [22,23] is usually measured by the clustering coefficient, which is expressed as the ratio of the actual number of edges of a node in the network with the surrounding nodes to the maximum of the theoretical number of edges, reflecting the degree of aggregation of the subway network sites and being more concerned with the local characteristics. The higher the clustering centrality, the more concentrated the stations in the area.…”