基于注意力机制的互特征融合旋转机械故障检测技术

Mutual feature fusion fault detection technique of rotating machinery based on attention mechanism

摘要:

旋转机制在生产生活中的应用愈加广泛。但旋转机械的存在应用环境较为复杂,生产环境恶劣,各部件相互影响,单一信号无法完整表现故障特征等问题。针对此问题,研究根据注意力机制构建卷积神经网络,采用多信号源进行数据提取,将不同信号特征相互融合构建旋转机械故障检测模型。实验结果表明,构建模型的故障分类准确率为99.92%,比第二优的算法高出1.89%,数据进行傅里叶变换后的检测精度平均提升了17.32%。由此可得,构建的故障检测模型能够有效提取并融合不同数据采集的故障特征,能够大幅提升旋转机械的故障检测精度,且将数据特征融合模块加入模型中能够有效减少单独计算的运行成本,提高运算速度。减少了因机械故障产生的生产安全事故,可以有效提升产品质量,提高企业的市场竞争力。

Abstract:

Rotation mechanism is more and more widely used in production and life. However, the application environment of rotating machinery is more complex, the production environment is harsh, the components affect each other, and a single signal can not complete the performance of fault characteristics. To solve this problem, a convolutional neural network was constructed according to the attention mechanism, multiple signal sources were used for data extraction, and different signal features were fused to build a rotating machinery fault detection model. The experimental results show that the fault classification accuracy of the constructed model is 99.92%, 1.89% higher than that of the second best algorithm, and the detection accuracy of the data after Fourier transform is improved by 17.32% on average. It can be concluded that the fault detection model constructed can effectively extract and fuse fault features of different data acquisition, which can greatly improve the fault detection accuracy of rotating machinery, and the addition of data feature fusion module to the model can effectively reduce the operating cost of separate calculation and improve the operation speed. It reduces the production safety accidents caused by mechanical failures, can effectively improve product quality and improve the market competitiveness of enterprises.