The installed capacity of renewable energy in power systems is rising rapidly in recent years due to environmental pressure. And as the main asset of mitigating renewable output fluctuations, energy storage (ES) also has been greatly developed with the increase of renewable capacity. To this end, the capacity planning of renewables and ESs has drawn much attention and many methods have been proposed. In the formulations of capacity planning problems, more detailed and complicated operation constraints mean more accurate planning results, which is the consensus of many researchers. However, to guarantee the problem formulation tractable, the actual multistage operation process of power system is not properly considered in existing planning methods. Therefore, in this paper, a new capacity expansion planning method for wind power and ESs is proposed considering the actual multistage operation process of power system. Specially, the hourly robust transmission constrained unit commitment (TCUC) and economic dispatch (ED) are involved in this planning method and thus, it could accurately evaluate the operational cost under certain planning decision. Besides, to guarantee the solvability and computational efficiency of planning method, a parallel horizon-splitting method is improved to solve the long-time-horizon hourly robust TCUC problem, and a genetic algorithm nested gradient descent method is established to accelerate the solving of planning problem. With the actual multistage operation process, a more reasonable expansion planning decision is obtained and the computational efficiency is greatly improved by the proposed acceleration algorithm. Numerical tests verify the efficacy of proposed method.
Read full abstract