As wildfires increasingly threaten power system infrastructure across North America and globally, grid operators face mounting pressure to make rapid, informed decisions under extreme uncertainty. A new research framework addresses this challenge by automating operational decision-making through advanced optimization techniques specifically designed for evolving wildfire scenarios.
The proposed system treats wildfire response as a dynamic, multi-stage problem where conditions change continuously. Rather than preparing a single operational plan, the framework generates adaptive decision trees that branch based on actual wildfire progression and system geography. This allows operators to adjust strategies—including transmission line switching, load shedding, and generator dispatch—as threats materialize or dissipate.
The methodology integrates both prevention and correction. Preventive actions reduce exposure before fires arrive; corrective actions minimize damage once threats emerge. The optimization objective balances three competing priorities: reducing wildfire-related risk, controlling operational costs, and minimizing customer blackouts.
A key innovation is the stochastic dual dynamic programming algorithm, which solves the optimization problem to global optimality despite the vast number of possible scenarios. Testing on the IEEE 30-bus and 300-bus standard test systems demonstrates the approach scales effectively to realistic grid sizes.
Results show the automated framework substantially outperforms conventional single-stage optimization. By making sequential, scenario-informed decisions rather than committing to one plan upfront, operators can better adapt to wildfire behavior, reduce unnecessary load shedding, and lower overall system costs.
The work carries significant practical implications for utilities in fire-prone regions, particularly in western North America, Australia, and Mediterranean climates. Integration with real-time wildfire forecasting and grid monitoring systems could enable operators to implement these strategies operationally, improving resilience when communities need it most.
