Fatigue damage accumulation needs to be monitored for extending the lifetime of horizontal axis wind turbines. To monitor accumulation of fatigue damage, this paper proposes a blade load and fatigue damage monitoring system consisting of physics models coupled with the Kalman filter (KF). The proposed monitoring system estimates bending moment distribution along the blade as a time series from 1 Hz data collected by a supervisory control and data acquisition (SCADA) system and from 20 Hz strain sensor data collected at the blade root. Static load is computed on the basis of blade element and momentum theory from SCADA data, whereas dynamic response of the blade is computed by assuming a mass–spring–damper system. The KF corrects the estimated moment distribution with moment measured at the blade root. The monitoring system is demonstrated and validated by installing it on a 2 MW floating downwind turbine and estimating strain and fatigue damage at the medium span position of the blades, where a strain sensor is installed. The validation results show that strain wave forms and fatigue damage estimated by the monitoring system with the KF are consistent with those computed from the strain measurement with sufficient accuracy for structural health monitoring.

需要监测疲劳损伤累积以延长水平轴风力涡轮机的使用寿命。为了监测疲劳损伤的累积，本文提出了一种由物理模型和卡尔曼滤波器（KF）组成的叶片载荷和疲劳损伤监测系统。所提出的监测系统根据监控和数据采集 (SCADA) 系统收集的 1 Hz 数据和在叶片根部收集的 20 Hz 应变传感器数据，将叶片上的弯矩分布估计为时间序列。静载荷是根据来自 SCADA 数据的叶片元素和动量理论计算的，而叶片的动态响应是通过假设质量-弹簧-阻尼器系统来计算的。KF 使用在叶根处测量的力矩来校正估计的力矩分布。通过将其安装在 2 MW 浮动顺风涡轮机上并估计安装了应变传感器的叶片中等跨度位置处的应变和疲劳损伤，对监控系统进行了演示和验证。验证结果表明，带有 KF 的监测系统估计的应变波形和疲劳损伤与根据应变测量计算的结果一致，具有足够的精度用于结构健康监测。