CAJ | 학술논문

考虑到风电功率爬坡事件在大规模、高集中度的风电发展模型下危害较大，为降低爬坡危害，维护电网的正常运行，提出了基于气象背景选取邻近点的爬坡事件多变量预测方法。首先，为提高风电功率预测的可靠性，充分考虑多气象因子的影响，通过对气象数据进行相空间重构建立了含多变量的基本预测模型。其次，考虑爬坡事件与气象背景间的联系，基于相关性分析确定不同气象背景下邻近点的选取准则。结合支持向量机分类模型，通过训练给出过去不同气象条件对应的邻近点选取机制，并指导选取预测所需的邻近点。最后，结合风电功率预测结果和爬坡事件检测方法，对实例数据进行爬坡仿真和预测，验证所提方法的有效性和可行性。
고필도풍전공솔파파사건재대규모、고집중도적풍전발전모형하위해교대，위강저파파위해，유호전망적정상운행，제출료기우기상배경선취린근점적파파사건다변량예측방법。수선，위제고풍전공솔예측적가고성，충분고필다기상인자적영향，통과대기상수거진행상공간중구건립료함다변량적기본예측모형。기차，고필파파사건여기상배경간적련계，기우상관성분석학정불동기상배경하린근점적선취준칙。결합지지향량궤분류모형，통과훈련급출과거불동기상조건대응적린근점선취궤제，병지도선취예측소수적린근점。최후，결합풍전공솔예측결과화파파사건검측방법，대실례수거진행파파방진화예측，험증소제방법적유효성화가행성。
With wind power developing inlarge-scale and high concentration in China, the harm brought by wind power ramp events is serious. In order to reduce this harm and keep safe operation of power system, this paper proposes a multi-variable prediction method of wind power ramps using meteorological condition to select adjacent points. The method makes use of meteorological data to improve performance of prediction. Firstly, it reconstructs phase space of meteorological factors and wind power series, and establishes basic multi-variable prediction model. Then, considering relationship between meteorological conditions and ramp events, the principle of selecting adjacent points under different meteorological conditions is given through correlation analysis. For purpose of guiding selection ofadjacent points of future predicted points, this paper utilizes support vector machine (SVM) classification model to train adjacent point selection mechanism on different meteorological conditions. Finally, combining method of wind power prediction and ramp detection, simulation and prediction of wind power ramp events on example data are completed. Through assessment analysis of the results, the proposed method is verified to be feasible and effective.