Structural Evolution and Resilience of Digital Economy Ecosystems: A Joint Temporal Exponential Random Graph Model...

Existing static analysis methods neglect the structural evolution of complex network topologies and the cascading failures induced by local load redistribution, leading to assessment biases in the analysis of the resilience of digital economy ecosystems. To accurately quantify the system's resilience threshold, this paper proposes a physical computation framework combining a joint Temporal Exponential Random Graph Model (TERGM) and an improved Motter-Lai algorithm (TERGM-ML). This framework utilizes Markov Chain Monte Carlo Maximum Likelihood Estimation (MCMC-MLE) to model endogenous structural effects and reconstruct the temporal evolution trajectory of the real network topology to overcome the limitations of static baselines. Subsequently, based on the centrality of nodes and their nonlinear physical capacity, a traffic redistribution rule dependent on the remaining capacity of neighbors is triggered when encountering a deliberate attack, tracking the entire process of system disintegration caused by local overload propagation. Multimodel comparative simulations show that, after introducing a dual mechanism of temporal evolution and dynamic reallocation, the critical node removal threshold that triggers a global transmission efficiency collapse in a deliberate attack scenario based on betweenness centrality is 12.41% ± 0.63%, which is significantly higher than the static scale-free baseline (7.85% ± 0.42%, p < 0.001).