多参数模型风电机组叶片结冰监测与预警研究

STUDY ON MUTI-PARAMETER MODEL OF WIND TURBINE BLADE ICING DETECTION AND WARNING

详细分析叶片结冰对风电机组运行性能和运行参数的影响,采用功率、叶轮转速和环境温度作为监测叶片结冰的变量。采用高斯过程回归分别建立功率模型和叶轮转速模型实现2个参数的实时监测。引入序贯概率比检验方法分析功率和叶轮转速模型的预测残差以发现2个参数在叶片结冰时的异常变化。当风电机组功率异常、叶轮转速异常且环境温度在0℃附近这3个条件同时满足时,发出叶片结冰预警。以云南某高原风场风电机组叶片结冰实际数据为例,验证本文方法的有效性。

This paper detailed analyzes the influences of the ice accretion on wind turbine performance and parameters. Output power,rotor speed and ambient temperature are selected as monitoring variables for blade ice accretion. Models are constructed using Gaussian Process Regression(GPR)to timely monitor output power and rotor speed parameters. Sequential Probability Ratio Test(SPRT)is introduced to analyze the predicting residual in order to detect the output power and rotor speed abnormalities. If three conditions including power abnormality,rotor speed abnormality,and ambient temperature near zero Celsius degree occur coincidence,blade icing alarm is triggered. The effectiveness of the proposed method is verified by taking the actual data of blade icing in a plateau wind farm in Yunnan as an example.