The integration of wind power requires the power system to be sufficiently flexible to accommodate its forecast errors. In the market clearing process, the scheduling of flexibility relies on the manner in which the wind power uncertainty is addressed in the unit commitment (UC) model. This paper presents a novel risk-based day-ahead unit commitment (RUC) model that considers the risks of the loss of load, wind curtailment and branch overflow caused by wind power uncertainty. These risks are formulated in detail using the probabilistic distributions of wind power probabilistic forecast and are considered in both the objective functions and the constraints. The RUC model is shown to be convex and is transformed into a mixed integer linear programming (MILP) problem using relaxation and piecewise linearization. The proposed RUC model is tested using a three-bus system and an IEEE RTS79 system with wind power integration. The results show that the model can dynamically schedule the spinning reserves and hold the transmission capacity margins according to the uncertainty of the wind power. A comparison between the results of the RUC, a deterministic UC and two scenario-based UC models shows that the risk modeling facilitates a strategic market clearing procedure with a reasonable computational expense.
Ieee Transactions on Power Systems, 2015, Vol 30, Issue 3, p. 1582-1592
Convex model; day-ahead market clearing; probabilistic forecast; risk-based unit commitment; wind power integration; Commerce; Forecasting; Integer programming; Integration; Probability distributions; Risk perception; Scheduling; Weather forecasting; Wind power; Convex modeling; Day ahead market; Probabilistic forecasts; Unit-commitment; Wind power integrations; Power markets; Components, Circuits, Devices and Systems; Power, Energy and Industry Applications; Equations; Generators; Load modeling; Mathematical model; Uncertainty; Wind forecasting; Wind power generation; Computational expense; Mixed-integer linear programming; Objective functions; Piecewise linearization; Probabilistic distribution; Transmission capacities