With the increasing penetration of distributed energy resources (DERs) in the active distribution network (ADN), how to enable joint planning of DERs under the uncertainty of distributed generations (DGs) has become a...
详细信息
With the increasing penetration of distributed energy resources (DERs) in the active distribution network (ADN), how to enable joint planning of DERs under the uncertainty of distributed generations (DGs) has become a challenging problem. This study establishes a two-stage joint planning model considering doubly-fed induction generator, photovoltaics (PVs) with the ancillary services of PV inverter, distributed energy storage systems and different types of controllable loads in the ADN. To address the uncertainties of DGs, a two-stage data-driven distributionally robust planning model is constructed. The proposed model is solved in a 'master and sub-problem' framework by column-and-constraint generation algorithm, where the master problem is to minimise the total cost and find the optimal planning decision under the worst probability distributions, and the sub-problem is to find the worst probability distribution of given uncertain scenarios. Besides, the original mixed-integer non-linear planning problem is converted into a mixed-integer second-order cone programming problem through second-order cone relaxation, Big-M and piecewise linearisation method. The numerical results based on 33-bus system verify the effectiveness of the proposed model.
This study proposes a novel resilience-directional robust dispatch (RRD) model for an islanded AC/DC hybrid microgrid (HMG). The inherent uncertainties on the source-load power and the occurrence of meteorological dis...
详细信息
This study proposes a novel resilience-directional robust dispatch (RRD) model for an islanded AC/DC hybrid microgrid (HMG). The inherent uncertainties on the source-load power and the occurrence of meteorological disasters are considered in this model. When a meteorological disaster strikes, the wind turbine (WT), photovoltaic (PV), and bidirectional converter of the HMG should be offline to ensure the stability of the HMG and the safety of these sensitive units. When affected by such double uncertainties, the output constraints of the WT, PV and load are bilinear but are linearised via big-M approach. The proposed RRD model manifests as a min-max-min tri-layer problem with mixed-integer recourse variables, which is difficult to solve directly. Therefore, a nested column-and-constraint generation algorithm is adopted to convert the tri-layer problem to a two-stage mixed-integer linear programming (MILP) model. The MILP problem is addressed by the commercial solver, thereby obtaining the minimal operating cost and establishing robust scheduling plans with the worst disaster scenario. The effectiveness and rationality of the proposed RRD model and its solution methodology are verified in numerical tests.
暂无评论