Data-Driven Transmission Defense Planning Against Extreme Weather Events

Ensuring the resilience of power system is the prerequisite for maintaining its normal operation after the occurrence of natural disasters. This paper proposes a data-driven transmission defense planning (DTDP) model to address the power system resilience issue against extreme weather events. DTDP enables the determination of the optimal resilience enhancement portfolio, including line hardening and construction. The target events of DTDP are selected according to survivability based severity indices. The component survivability and post-disaster system failure state are treated as random variables in two separate levels, and the selected historical data are used to construct the confidence sets of their ambiguous probability distribution. Thus, the man-made uncertainty budget can be avoided and the expected economic loss under the worst-case distribution is minimized. Moreover, a sparsification method of system failure states is designed to boost computational efficiency. The effectiveness of DTDP and its solution method are verified on a six-bus test system and RTS-79 system. The results demonstrate that the DTDP can obtain a better defense plan compared to other optimization techniques.