Bayesian Updating of Pole-Cable Systems for Damage Quantification after Sequential Extreme Events

Abstract

Strong earthquakes can significantly damage electrical distribution infrastructure, causing prolonged outages and challenging recovery. Sequential extreme events, such as earthquake mainshocks followed by aftershocks, further complicate recovery by requiring repeated damage assessments and repairs. Utilities need real-time tools that combine physics-based modeling with power-infrastructure monitoring to assess damage and prioritize repairs during disaster response. This paper presents a Bayesian inference approach that updates the nonlinear structural properties of pole-cable systems after each hazard sequence using time-history response data. We demonstrate the method using a sequence of earthquakes with no intermediate repairs. First, we developed a high-fidelity finite element model of a cable-connected pole system that captures material and geometric nonlinearities and the progressive damage that develops under sequential earthquake loading. Second, we implement Bayesian inference to sequentially update the nonlinear parameters of the cable-connected pole model using recorded time-history data after each earthquake. We introduce an energy-based log-likelihood function for Bayesian inference of pole properties and employ the Teleport Markov Chain Monte Carlo algorithm to improve exploration and mixing of the chains in the complex multidimensional space. Finally, we track the degradation of power pole strength that may trigger repairs. The proposed framework can provide utilities with real-time information on damage states of individual poles, supporting prioritization of inspections and repairs during sequential earthquake events.