Distributionally robust optimal dispatch in the power system with high penetration of wind power based on net load fluctuation data

The power system with high penetration of wind power faces a great challenge for system dispatch due to the high volatility and intermittency of the wind power. This work proposes a day-ahead optimal dispatch model which is formulated for a power system with thermal power, hydropower, and controllable load as dispatchable resources. According to the anti-peak regulation, the system dynamic power regulation margin model considering adjacent time periods is established to address the uncertainty in the fluctuation rate of net load power, and to sufficiently use the dispatchable resources to reduce wind curtailment. However, in some circumstances, curtailing wind has to be considered to ensure secure operation of the system and maintain the economic goal from the total cost point of view. The risk of curtailing wind is formulated using conditional value at risk (CVaR), and is minimized as part of the total operating cost. Another objective function of the proposed dispatch model is to maximize the power regulation margin. Uncertainty in the fluctuation rate of net load power is modelled for different weather conditions using moment-based ambiguity set. The moment information is obtained from large amount of historical data using clustering methodologies. The proposed optimal dispatch model with uncertainty and CVaR formulation is reformulated and solved under the distributionally robust conditional value at risk (DRCVaR) framework. The model is transformed into a semi-definite programming problem through the duality theory and can be solved efficiently by commercial solvers. Simulation results show that the proposed dispatch model can effectively strengthen wind power absorption, ensure secure operation, and improve the robustness of the dispatch strategy facing the uncertainty from the wind power.

Authors:Bowen Chen, Emmanuel Ackom, Hongming Yang, Huan Zhao, Junpeng Liu, Rui Liang, Sheng Xiang, Yuan Yuan
Published year:2022
Content type:Journal article
Orbit ID:73dba155-4d14-47d1-8829-223a690104af
Is current:Current