Abstract:

In order to clarify the correlation of eVTOL risk factors, and explore their impact on risk prevention and control in the UAM ecosystem, complex network theory was used to establish a risk evolution model. Based on the UAV accident database at home and abroad and the statistics of general aviation accidents, combined with the operation characteristics of eVTOL in urban low-altitude scenes, 35 types of risk factors and 10 types of dangerous events were identified from the perspective of human-machine-environment. Gephi software was used to construct the network model, and the key nodes were evaluated comprehensively by the node degree, proximity centrality, internode centrality and PageRank(PR) algorithm. The key edges were evaluated by the internode number, so as to determine the key risk propagation path. In order to reduce the system risk, the measures to reduce the chain breaking disaster were proposed, and the system safety after chain breaking control was measured by network efficiency index. The results show that there are strong correlations among the eVTOL risk factors in the UAM ecosystem, and there are eight key risk transmission chains. The system safety is improved by 4.74%, 16.21% and 18.10% by blocking key human factors, key system technical failure factors and key intermediate dangerous events, respectively.

Key words: complex network, urban air mobility (UAM), electric vertical takeoff and landing (eVTOL), operational risk, broken chain control