Invulnerability Simulation of Urban Agglomeration Traffic Network Cascading Failure Under Random Attack Strategy

Aiming at the cascading failure problem in urban agglomeration traffic network, this paper proposed a new cascading failure model. In this model, we divided the node states into normal, pause, and failure with the judgment criterion of the degree of node overload. For the pause node we adopted the flow distribution rule of residual capacity priority, and for the failure node, the flow distribution was based on the node distance and degree of connectivity. Meanwhile, this paper also considered the enduring ability between nodes or edges in the load transfer process. We studied the effects of overload capacity parameter which means the node redundancy capacity, passenger flow adjustment which means the initial network load, and the impact of different attack strategies on urban agglomeration traffic network invulnerability with the simulation of actual passenger flow in Hu-Bao-E urban agglomeration. The results show that the overload capacity parameter has a threshold of 1.2 with the node attacks, so the invulnerability would not be improved when we increase the value of overload capacity parameter. The influence of passenger flow adjustment parameter on network invulnerability shows different characteristics under different attack patterns. Compared with the attack edge, the attack node has a more significant impact on traffic network invulnerability, because node bears large passenger flow, and it will easily be cascading failure when it is attacked. And in the attack process, it shows the cliff-like decline phenomenon of correction relative scale of maximum connected subgraph, and the urban agglomeration traffic network has an extremely strong scale-free characteristic.