Two-phase gating attention time sequence classification method and system based on two-way jump storage

• Main Authors: MORIE, LIU MIN, QIN LE, LUO YONG, WU FUYING, PENG YALI, YI YUGEN, HUANG XIAOMEI
• Format: Patent
• Language: Chinese; English
• Published: 05.04.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention provides a two-phase gating attention time sequence classification method and system based on two-way jump storage, which are applied to machine fault detection, and the method comprises the following steps: obtaining a multivariate time sequence data set of a target machine, respectively extracting multivariate time sequence data input in a time sequence and an inverted sequence, inputting the multivariate time sequence data into a two-way jump storage encoder, performing jump connection of different time steps to obtain echo state representations of different time scales, respectively inputting a gating circulation attention layer and a multi-scale gating attention layer in a dual-stage gating attention learning device to obtain a first feature representation and a second feature representation, performing fusion to obtain a fusion feature, and performing multi-scale gating attention learning on the fusion feature to obtain the echo state representations of different time scales. And inputting