Zhigang Li team's condition speeds energy storage dispatch 10x with exact relaxation
Researchers prove when you can ditch MIP without losing optimality, slashing compute time.
Deep Dive
Shan Liu, Zhigang Li, and Ye Guo propose a necessary and sufficient condition for exact relaxation of the complementarity constraint in energy storage power dispatch. They derive the relaxation gap and show that simultaneous charging/discharging does not always change the optimal value from the mixed-integer programming model. In those cases, a feasible solution without simultaneous charging/discharging can be recovered without losing optimality. They also introduce a two-stage relaxation-and-recovery algorithm to improve computational efficiency.
Key Points
- MIP constraint for ESS dispatch is replaced with a relaxation condition that guarantees exactness (same optimal value).
- The two-stage algorithm first relaxes the complementarity constraint, then recovers a feasible SCD-free solution when needed.
- Computational speed improves dramatically—avoiding MIP's exponential runtimes while preserving objective value.
Why It Matters
Faster, exact power dispatch means cheaper grid operations and better integration of renewables at scale.