基于声发射技术的风电叶片复合材料损伤模式识别

Damage pattern recognition of wind turbine blade composite material based on acoustic emission technology

为研究风电叶片玻璃纤维复合材料在疲劳工况下的损伤模式,文章基于声发射技术提出了一种主成分聚类分析和BP神经网络相结合的材料损伤识别模型。首先,采集损伤声发射信号,并提取相关参数进行分析,对不同疲劳损伤进行分类;其次,对数据进行主成分分析,以降低噪声信号,去掉冗余信息;再次,对主成分进行聚类分析,将样本分簇并找出各簇与损伤之间的对应关系;最后,基于BP神经网络建立损伤识别模型,并基于试验数据对识别网络进行测试训练。训练结果表明,识别模型对3种未知类型疲劳损伤的识别率均高于90%,对未知损伤具有较好的识别能力。

In order to study the damage mode of wind turbine blade glass fiber composite materials in fatigue conditions, a material damage identification model based on the combination of principal component cluster analysis and BP neural network is proposed. First, collect damage acoustic emission signals and extract relevant parameters for analysis to classify different fatigue damage. Then, perform principal component analysis on the data to reduce noise signals and remove redundant information,and then perform cluster analysis on principal components to cluster the samples and find the correspondence between each cluster and the damage. Finally, a damage recognition model is established based on BP neural network, and the recognition network is tested and trained based on experimental data. The results show that the recognition rate of the three unknown types of fatigue damage by the recognition model is higher than 90%. The damage recognition model proposed in this paper has a good ability to recognize unknown damage.