【摘要】 论文提出一种基于双向LSTM的气动回路故障诊断方法，使用AMESim软件建立某生产线的气动回路仿真模型，模拟气动回路的单重故障和多重故障，记录仿真数据，制作该气动回路的故障诊断数据集，在Matlab环境下建立双向LSTM网络架构进行气动回路故障诊断实验与分析，结果表明双向LSTM模型在多重故障识别的正确率高于LSTM模型以及传统的诊断方法。更多还原

【Abstract】 This paper presents a fault diagnosis method of pneumatic circuit based on bidirectional LSTM. AMESim software is used to establish the simulation model of pneumatic circuit of a production line,the single fault and multiple fault of pneumatic circuit are simulated,the simulation data is recorded,a fault diagnosis data set of the pneumatic circuit is made,and the bidirectional LSTM network architecture is established in Matlab environment for the pneumatic circuit fault diagnosis experiment and analysis. The results show that the accuracy of bidirectional LSTM model is higher than that of LSTM model and traditional diagnosis methods.更多还原