基于CNN-LSTM-Attention模型的湿喷台车泵送系统堵管故障预测方法

王开松,魏一鸣,唐威,郭旭华,李朝阳,邹俊

Fault prediction method of pipeline blockage in wet spray trolley pumping system based on CNN-LSTM-Attention model

Kaisong WANG,Yiming WEI,Wei TANG,Xuhua GUO,Zhaoyang LI,Jun ZOU

表2 CNN-LSTM-Attention模型在不同预测长度下的预测性能

Table 2 Prediction performance of CNN-LSTM-Attention model under different prediction lengths

预测长度 ERMS EMA R2 12 1.49 0.58 0.946 24 1.89 0.81 0.912 36 2.02 0.94 0.905 48 2.24 1.11 0.878