A heuristic method to identify optimum seismic retrofit strategies for critical infrastructure systems

Chuang Liu,Min Ouyang,Naiyu Wang,Zijun Mao,Xiaolin Xu

DOI: https://doi.org/10.1111/mice.12647

2021-01-01

Abstract:Retrofitting critical components of a critical infrastructure system to improve its seismic performance has been considered as the most frequently used mitigation strategy in both literature and practice. This article mainly studies this mitigation strategy and formulates the seismic retrofit optimization problem for critical infrastructure systems under a limited retrofit budget in a general form, and then proposes a heuristic method to solve the problem efficiently in terms of the optimality gap and the computational cost. The proposed method mainly includes three steps: (1) generates a limited number of component damage scenarios to reformulate the problem as an approximated model; (2) adopts a component importance-based method to reduce the solution space and applies the integer L-shaped method to solve the approximated model; (3) employs the sample average approximation method to enhance the solution quality. To demonstrate the performance of the proposed method, it is applied to identify the optimal retrofit strategies for the Shelby power transmission system, the IEEE 14-bus test system, and the IEEE 118-bus test system and also compared with several existing methods. Results show that the proposed method is significantly more efficient than those existing methods. For the IEEE 14-bus test system, the proposed method gets almost exact solutions, with errors less than 0.29%; for the other two systems, it returns the best solutions among all methods under various retrofit budgets.