论文部分内容阅读
为克服规模化风电场中MCR型SVC控制性能的局限性,避免故障切除后无功过补偿,本文提出了基于风电场动态电压安全决策树体系的SVC紧急控制策略。分别基于决策树中的回归树算法与分类树算法构建风机过电压预测回归树以及风机脱网分类树,形成风电场动态电压安全决策树体系。该决策树体系依据在线信息进行数据挖掘,对风机脱网状况与电压越限情况进行快速预判,并根据预测结果,采取合理的电容器退出措施,避免无功过补偿。算例分析表明,所提策略不仅能为电网及风电场运行人员提供风机脱网风险信息与决策参考,还能够降低故障切除后风机因过电压而脱网的风险。
In order to overcome the limitations of MCR SVC in large-scale wind farms and avoid reactive power compensation after fault cut, an SVC emergency control strategy based on dynamic voltage safety decision tree in wind farm is proposed in this paper. Based on the regression tree algorithm and the classification tree algorithm in the decision tree respectively, the wind turbine overvoltage prediction regression tree and the wind turbine off-net classification tree are constructed to form the dynamic voltage safety decision tree system of the wind farm. The decision tree system uses data mining based on online information to quickly predict fan off-grid condition and voltage over-limit condition. Based on the predicted results, reasonable capacitor withdrawal measures are adopted to avoid reactive over-compensation. The case study shows that the proposed strategy can not only provide information on wind turbine risk and decision-making, but also reduce the risk of fan disconnecting after over-voltage.