MDP based Decision Support for Earthquake Damaged Distribution System Restoration

TL;DR

This paper introduces an MDP-based decision support framework for restoring electricity in damaged distribution systems after earthquakes, aiming to maximize customer service and minimize restoration time.

Contribution

It develops a novel MDP-based method that incorporates earthquake damage probabilities to optimize power system restoration strategies.

Findings

Effective restoration schemes identified using the MDP approach.

Quantitative assessment of damage probabilities based on PGA data.

Enhanced decision-making for power system resilience post-earthquake.

Abstract

As the society becomes more dependent on the presence of electricity, the resilience of the power systems gains more importance. This paper develops a decision support method for distribution system operators to restore electricity after an earthquake to the maximum number of customers in the minimum expected duration. The proposed method employs Markov Decision Process (MDP) to determine the optimal restoration scheme. In order to determine the probability of the field component damage due to the earthquake, the Probability of Failure ($P_f$) of structures are calculated using the Peak Ground Acceleration (PGA) values recorded by observatories and earthquake research centers during the earthquake.