深度学习(deep learning,DL)应用于推动智能故障诊断的发展,带来了显著的性能提升。然而,现有方法大多无法捕捉机械设备的时间信息和全局特征,无法收集到足够的故障信息。同时,由于运行环境复杂恶劣,单源故障诊断方法难以稳定、广泛地...深度学习(deep learning,DL)应用于推动智能故障诊断的发展,带来了显著的性能提升。然而,现有方法大多无法捕捉机械设备的时间信息和全局特征,无法收集到足够的故障信息。同时,由于运行环境复杂恶劣,单源故障诊断方法难以稳定、广泛地提取故障特征。因此,提出一种基于多源信息融合(multi-source information fusion,MSIF)的连续小波和TransXNet三相异步电机故障诊断新方法,通过提取和整合丰富的特征来提高诊断性能稳定性。首先,搭建三相异步电机故障实验平台,使用加速度传感器与电流传感器采集电机多种工况下的振动信号与电流信号,获得多源信号;其次,提出一种新的轻量级混合网络模块:双动态令牌混合器(dual dynamic token mixer,D-Mixer),其可以动态地利用全局和局部信息,同时注入大的感受野和强大的归纳偏差,而不牺牲输入依赖性,提高了特征提取能力。提出多尺度前馈网络(multi-scale feed-forward network,MS-FFN),在前馈网络中进行多尺度特征聚合。通过交替使用D-Mixer和MS-FFN,构建一种新型的混合CNN-Transformer网络:TransXNet;然后,利用连续小波变换将多源信号进行时频变换,提出数据级融合策略获得多源信息图,将多源信息图输入到TransXNet中进行特征分割以及聚合以完成特征提取,以训练并验证所提出的TransXNet有效性;最后,使用多源测试样本来验证所提出方法的诊断性能。结果表明,基于TransXNet强大的特征提取能力,识别准确率达到100%。通过对比调整兰德指数、归一化互信息、F1分数和准确率4个评价指标以及抗噪性分析,得出所提方法优于目前故障诊断领域最先进的方法(state of the art,SOTA),在故障诊断领域具有很好的前景。展开更多
For the characteristics of the continuous stirred-tank reactor(CSTR) with coil and jacket cooling system,a CSTR temperature dual control solution based on the analysis of the CSTR exothermic reaction control character...For the characteristics of the continuous stirred-tank reactor(CSTR) with coil and jacket cooling system,a CSTR temperature dual control solution based on the analysis of the CSTR exothermic reaction control characteristic was proposed for an organic material polymerization production.The control solution has passive fault-tolerant ability for the jacket cooling water cutting off fault and active fault-tolerant potential for the coil cooling water cutting off fault,and it has good control ability,high saving energy and reducing consumption performance.Fault detection and diagnosis and fault-tolerant control strategy are designed for the coil cooling fault to achieve the active fault-tolerant control function.The CSTR temperature dual control,process fault detection and diagnosis and active fault-tolerant control were full integrated into the CSTR temperature fault-tolerant control system,which achieve fault tolerance control of CSTR temperature for any severe malfunction of jacket cooling or coil cooling cutting off,and the security for CSTR exothermic reaction is improved.Finally,the effectiveness of this system was validated by semi-physical simulation experiment.展开更多
文摘深度学习(deep learning,DL)应用于推动智能故障诊断的发展,带来了显著的性能提升。然而,现有方法大多无法捕捉机械设备的时间信息和全局特征,无法收集到足够的故障信息。同时,由于运行环境复杂恶劣,单源故障诊断方法难以稳定、广泛地提取故障特征。因此,提出一种基于多源信息融合(multi-source information fusion,MSIF)的连续小波和TransXNet三相异步电机故障诊断新方法,通过提取和整合丰富的特征来提高诊断性能稳定性。首先,搭建三相异步电机故障实验平台,使用加速度传感器与电流传感器采集电机多种工况下的振动信号与电流信号,获得多源信号;其次,提出一种新的轻量级混合网络模块:双动态令牌混合器(dual dynamic token mixer,D-Mixer),其可以动态地利用全局和局部信息,同时注入大的感受野和强大的归纳偏差,而不牺牲输入依赖性,提高了特征提取能力。提出多尺度前馈网络(multi-scale feed-forward network,MS-FFN),在前馈网络中进行多尺度特征聚合。通过交替使用D-Mixer和MS-FFN,构建一种新型的混合CNN-Transformer网络:TransXNet;然后,利用连续小波变换将多源信号进行时频变换,提出数据级融合策略获得多源信息图,将多源信息图输入到TransXNet中进行特征分割以及聚合以完成特征提取,以训练并验证所提出的TransXNet有效性;最后,使用多源测试样本来验证所提出方法的诊断性能。结果表明,基于TransXNet强大的特征提取能力,识别准确率达到100%。通过对比调整兰德指数、归一化互信息、F1分数和准确率4个评价指标以及抗噪性分析,得出所提方法优于目前故障诊断领域最先进的方法(state of the art,SOTA),在故障诊断领域具有很好的前景。
基金Project(2013JM8024)Supported by Natural Science Basic Research Plan in Shaanxi Province of China
文摘For the characteristics of the continuous stirred-tank reactor(CSTR) with coil and jacket cooling system,a CSTR temperature dual control solution based on the analysis of the CSTR exothermic reaction control characteristic was proposed for an organic material polymerization production.The control solution has passive fault-tolerant ability for the jacket cooling water cutting off fault and active fault-tolerant potential for the coil cooling water cutting off fault,and it has good control ability,high saving energy and reducing consumption performance.Fault detection and diagnosis and fault-tolerant control strategy are designed for the coil cooling fault to achieve the active fault-tolerant control function.The CSTR temperature dual control,process fault detection and diagnosis and active fault-tolerant control were full integrated into the CSTR temperature fault-tolerant control system,which achieve fault tolerance control of CSTR temperature for any severe malfunction of jacket cooling or coil cooling cutting off,and the security for CSTR exothermic reaction is improved.Finally,the effectiveness of this system was validated by semi-physical simulation experiment.
