Research on risk propagation method of multimodal transport network under uncertainty

Taking multimodal transport network as the research object, a quantitative method of risk propagation based on improved percolation theory is proposed. This paper analyzes the applicability and limitations of the percolation theory for this problem, and improves such four aspects of nodes and edges as the state, the initial load, the percolation probability and the evaluation index of percolation effect, so that they are universal to the risk propagation of the transport network. Taking Sichuan–Tibet​ region as an example for empirical analysis, this paper simulates and analyzes the risk propagation law of the multimodal transport network under different attack types and load preferences. The results show that the risk propagation effect in the multimodal transport network will become stronger with the increase of attack scale and attack intensity, and it will show an exponential growth trend with the increase of attack scale. When other conditions are the same, the order of strength of the three kinds of attack law according to the risk propagation effect is: the intentional attack based on the loads > the random attack > the intentional attack based on the failure probability; the risk propagation effect of the load with preferences is stronger than that without preferences. Therefore, managers can control the risk of multimodal transport network in Sichuan and Tibet from the perspective of controlling attack types and balancing load preference. The research improves the method of risk propagation simulation, makes up for the shortcomings of existing research which neglects the subjectivity of load and the dynamics of risk propagation, and makes it closer to the reality, which is of strong practical value and theoretical significance. •We discusses the limitations of the percolation theory application in the multimodal transport network, and proposes a universal improvement method.•This paper proposes that the subjectivity of load has a certain influence on the risk propagation effect in the network. In order to quantify the subjectivity of load, we use the FMEA, the triangular fuzzy number and the prospect theory to improve the percolation probability, and integrate the subjectivity of load into the redistribution process of load.•This paper proposes the method of combining the static index and the dynamic index to measure the percolation effect, so as to explore the dynamics of risk transmission.•This paper makes an empirical analysis of the multimodal