大数据驱动下的风电机组工作状态决策树判别方法

Decision Tree Discrimination Method for Wind Turbine Working State Driven by Big Data

风电机组所处工作状态的准确判别,对于掌握风电机组是否安全稳定高效运行具有重要的价值和意义。基于北方某风电场SCADA系统历史大数据,利用统计原理对风电机组特征参数进行了选取,在此基础上建立了一种判别风电机组工作状态的决策树(Decision Tree Algorithm)算法模型,并对模型参数进行了网格搜索优化,同时使用了新的历史数据对模型泛化能力做了检验,递归计算了风电机组不同状态下的信息增益和信息熵。依据风场历史数据对风电机组6种典型工作状态模式做了判别验证,结果表明该方法能够准确有效地判别给出风电机组工作状态模式,准确率平均达到0.989,从而验证了此方法的可行性和有效性。

Accurate identification of the working state of wind turbines is of great value and significance to know whether they are safe,stable and efficient.Based on the historical big data of a SCADA system in a northern wind farm,the statistical principle was used to select the characteristic parameters of the wind turbine.On this basis,a decision tree algorithm model for determining the operating status of the wind turbine was established,whose parameters were optimized by grid search.At the same time,new historical data were used to test the generalization ability of the model.The information gain and information entropy of the wind turbine in different states were recursively calculated.According to the historical data of the wind field,the six typical working modes of the wind turbine were discriminated and verified.The results show that the method in this paper can accurately and effectively identify the working mode of the wind turbine,with an average accuracy of 0.989,which verifies the feasibility and effectiveness of the method proposed in this paper.