大型风力发电机组运行稳定性劣化趋势预测方法研究

The trend of running state deterioration of large-scale wind turbine has a direct impact on the security, efficiency and long life operation of the system. If the state evolution form fault appearance to fault manifestation macroscopically can be revealed by trend prediction of running state deterioration, it will provide scientific means to implement predictive maintenance, avoid occurrence of fatal accidents. Then a variety of deterioration models of large-scale wind turbine were established，and the feature extracting method based on higher-order cumulant diagonal slice was studied. Results showed that the proposed state assessment method can describe the evolution track of stability deterioration accurately. An evaluation method using sensitivity property, trend property, difference property and consistency property was proposed, Experiment data of varying degrees of deterioration under various types of deterioration were carried out to validate the proposed method, and results showed that feature extracting method can separate the deterioration characteristic from non-degradation characteristics. The Hidden Markov model was applied to predict the trend of the deterioration’s syate.Real state vibration data were collected on the industrial scene and evaluation method research on the stability of operation. The results showed that the proposed state assessment method can predict deterioration trend and describe the evolution track of stability deterioration accurately. It also provided theoretical basis and scientific methods for the safety，stability，reliability of complex mechanical and electrical equipment such as the large-scale wind turbine，at the same time it can reduce the cost of the total life cycle of the equipment .

大型风力发电机组运行稳定性劣化趋势直接影响系统能否安全、高效和长寿命运行。若能够对运行稳定性劣化趋势进行有效预测，揭示劣化发生、发展直至恶化的演变过程，则有利于实施预知维护，避免恶性事故发生。本研究通过建立风力发电机组多种劣化类型模型，研究高阶累积量劣化特征提取方法，准确描述稳定运行状态劣化为非稳定运行状态的劣化动态演化机理，提出基于敏感性、趋势性、差异性、一致性判断特征提取方法的趋势预测适用性的方法，解决变工况、非平稳运行状态下，劣化特征与变工况等非劣化特征耦合难以分离，劣化趋势预测中特征提取方法的选择缺少理论依据的问题。应用隐马尔可夫模型对劣化状态进行趋势预测研究。开展工业现场风力发电机组实验验证及运行稳定性劣化状态评价方法的实验验证，对风力发电机组实际运行状态进行劣化趋势预测和有效评价，为提高大型风力发电机组复杂机电设备的安全性、稳定性以及降低全寿命周期费用提供理论依据和科学手段。

大型风力发电机组运行稳定性劣化趋势直接影响系统能否安全、高效和长寿命运行。若能够对运行稳定性劣化趋势进行有效预测，揭示劣化发生、发展直至恶化的演变过程，则有利于实施预知维护，避免恶性事故发生。本研究准确描述了稳定运行状态劣化和非稳定运行状态的劣化动态演化机理，提出了全矢频带能量谱故障诊断方法，解决了运行稳定性劣化进程中早期劣化特征信息微弱和稳定性劣化特征被变工况等非劣化信息淹没的问题。提出一种谱峭度和阶比跟踪相结合的故障特征提取方法，解决了变工况、非平稳运行状态下，劣化特征与变工况等非劣化特征耦合难以分离，劣化趋势预测中特征提取方法的选择缺少理论依据的问题。 提出了大型风力机整机动响应计算方法，运用模态叠加法对风力机的动态响应进行分析, 分析了影响风力机运行稳定性劣化趋势的因素。对风力机行星齿轮传动系统动态特性和齿轮劣化故障诊断研究，提出了利用经验模式分解(EMD)能量分布作为故障特征向量，灰色相似关联度作为故障模式识别算法的大型风力发电机齿轮劣化故障诊断方法，基于EWT-MDS和FIR分解的风力机轴承劣化趋势识别及故障诊断方法。对风力机叶片摆振运动表面位移与层间断裂韧性响应研究，确定柔性叶片摆振运动冲击对层间断裂劣化影响的趋势，提出了风力机叶片运行稳定性劣化趋势预测方法。提出一种基于布拉格光纤光栅（FBG）动态监测的风力机组故障诊断技术，构建了齿轮应变光纤光栅动态测量系统，对风力机组行星齿轮传动实际运行状态进行劣化趋势诊断验证和有效评价，证明了风力机传动系统运行稳定性劣化趋势预测方法的正确性，为提高大型风力发电机组复杂机电设备的安全性、运稳定性预测提供了理论依据和科学手段。通过项目的完成，发表科研核心论文29篇，其中SCI/EI收录检索14篇，获得省部级科技进步二等奖1项、三等奖1项，获得省自然科学优秀学术论文一等奖1项、三等奖1项。获得国家软件著作权登记2项，发明专利1项，专著1部待出版。培养博士后2人，培养博士3人，培养硕士研究生13名。