Optimal Reserve Allocation With Simulation-Driven Frequency Dynamic Constraint: A Distributionally Robust Approach

This brief proposes a simulation-driven framework that exploits the Monte Carlo simulation and the linear regression method to fit the implicit relationship between the integral frequency deviation and the frequency reserve allocations during the primary frequency response process of power systems. The fitted model is incorporated into a distributionally robust economic dispatch model as frequency dynamic constraints to optimally schedule the system while hedging against the uncertain regression residue. Numerical studies on the IEEE 39-bus system validate the effectiveness in fitting the relationship and dispatching frequency reserves.