摘要
风电机组风速计出现故障的概率较高,对其进行实时监测并及时发现其故障有重要意义。由于相邻多台风电机组的运行工况和风速计测量值的相关性很强,提出了基于相邻风机相关性模型的风速计监测方法。采用粒子群神经网络算法对相邻的多台风电机组风速计正常测量数据进行处理,建立相关性模型,将风速计实时测量风速作为模型的输入,当某台机组的风速计出现测量异常时,其与其他相邻机组风速计之间原有的相关性被破坏,相关性模型对该机组风速的预测残差将会显著增大,预示该风速计出现故障,据此能够实现风电机组风速计状态的实时监测。某风电场的实际运行数据验证了该方法的有效性。
It has important significance for real-time monitoring of wind turbine anemometer due to its high failure probability.This paper proposed an anemometer monitoring method based on adjacent wind turbines correlation model for the high correlation of operation conditions and measured values of anemometers among adjacent wind turbines. This paper processed the normal measured data of anemometers of adjacent wind turbines using particle swarm optimization( PSO) algorithm and built a correlation model. In this model,the real-time measured wind velocities of anemometers are considered as the inputs. When a certain wind turbine anemometer is abnormal,the correlation among adjacent wind turbines is destroyed and the predicted residual of the certain wind turbine anemometer will increase significantly,which indicates that anemometer may have a fault. Hence this method could realize the real-time monitoring of anemometer.The practical operation data of a certain wind farm verified the effectiveness of this method.
出处
《华电技术》
CAS
2016年第1期72-75,79,共4页
HUADIAN TECHNOLOGY
