摘要
为实现风电机组长期服役条件下功率输出特性劣化的监测与评估,考虑在同一场址条件下,地理位置相近,系统配置相同的风电机组外部环境条件具有相似性以及功率输出特性呈现强相关性,基于改进的非线性状态估计技术,采用改进欧式距离算法,建立风电机组功率特性劣化监测模型。结合某海上风电场风电机组簇实际运行数据,对该劣化监测方法和模型进行了测试。研究结果表明:基于改进非线性状态估计模型的风电机组功率特性劣化监测方法,正常情况下功率输出实际值与预测值的残差在0.5%以内;当风电机组功率输出特性出现劣化时,残差会出现异常并逐渐加大。提出的基于改进非线性状态估计的风电机组功率特性劣化监测方法,对于风电机组预测性运维以及优化技改具有指导意义。
To evaluate and monitor the wind turbine power performance degradation due to continuous increase of operation time in service,considering the similarity of the external environmental conditions and strong correlation of the power output characteristics of wind turbines with similar geographical locations and same configurations under the same site conditions,a wind turbine power performance degradation monitoring model,which is based on an improved nonlinear state estimate model and improved Euclidean distance method is established.Based on the operation data of all the wind turbine clusters from an offshore wind farm,the wind turbine power performance degradation method and model is tested and verified.The result shows that the residual error between the actual power value and the predicted power value is within 0.5%under normal conditions.Meanwhile,the residual error will be abnormal and gradually increase when the wind turbine power performance is degraded and so as to realize the power performance degradation timely and effectively.The power performance degradation monitoring method proposed,based on improved nonlinear state estimation,has practical guiding significance for the intelligent predictive operation and maintenance of wind turbines and also the optimization of their technical innovation.
作者
付德义
孔令行
FU Deyi;KONG Lingxing(State Key Laboratory of Operation and Control of Renewable Energy & Storage Systems, China Electric Power Research Institute, Beijing 100192, China)
出处
《兵器装备工程学报》
CSCD
北大核心
2021年第11期215-221,共7页
Journal of Ordnance Equipment Engineering
基金
国家重点研发计划项目(2018YFB0904000)课题5(2018YFB0904005)。
关键词
风电机组
功率特性
非线性状态估计
劣化监测
wind turbine
power performance
nonlinear state estimate
degradation monitoring
