【摘要】 针对现有风力机叶片两阶段退化可靠性评估方法的局限性,即忽略失效阈值随机性和退化过程变点准确性,提出一种新方法。该方法基于非线性维纳(Wiener)退化过程,考虑失效阈值随机性对不同阶段的影响。引入修正的赤池信息准则(AICc)确定最优变点位置,并进行区间估计。基于变点位置采用极大似然估计法确定两阶段漂移系数和扩散系数的估计值,然后通过疲劳裂纹扩展仿真实验得到相关数据,建立叶片退化的可靠性模型。同时,为验证模型预测的准确性,利用自适应算法更新后的风力机叶片数据进行实例分析。结果表明,失效阈值的随机性和变点准确性对退化建模具有显著影响,能够有效提高可靠性评估的准确性。更多还原

【Abstract】 To address the limitations of existing two-stage degradation reliability assessment methods for wind turbine blades, specifically the neglect of failure threshold randomness and the accuracy of changepoint detection in the degradation process, a novel approach is proposed. This method is based on a nonlinear Wiener degradation process and considers the impact of random failure thresholds across different stages. The corrected akaike information criterion(AICc) is introduced to determine the optimal changepoint location, followed by interval estimation. Maximum likelihood estimation is employed to determine the drift and diffusion coefficients for the two stages based on the identified changepoint. Subsequently, a reliability model for blade degradation is established using data obtained from fatigue crack propagation simulation experiments. To validate the accuracy of the model predictions, an empirical analysis is conducted using updated wind turbine blade data via an adaptive algorithm. The results demonstrate that accounting for failure threshold randomness and changepoint accuracy significantly impacts degradation modeling, thereby enhancing the precision of reliability assessments.更多还原