Motor equipment residual life prediction method based on multi-channel information refining convolution

• Main Authors: XUE SHUQIANG, YAN RUNSHENG, YANG KAI, XIE CHENG, LIU GUOPENG, XU LIANBING, HU WENFANG, JIA YUNHONG, BAO WENLIANG, GAO XUBIN, KANG YONGLING, DU CHUNHUI, HUO YAOBIN, GUO YONGHUI, YIN DONGXU
• Format: Patent
• Language: Chinese; English
• Published: 13.08.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention belongs to the technical field of residual life prediction of motor equipment, and solves the problem that the accuracy and generalization of residual life prediction of the motor equipment are seriously influenced as important degradation information of the motor equipment is difficult to effectively extract by a current method. The invention provides a motor equipment residual life prediction method based on multi-channel information refining convolution. The method comprises the following steps: acquiring original monitoring signals of motor equipment under different working conditions; constructing a degeneration behavior learning sub-network to obtain global advanced representation; constructing a residual life prediction sub-network to obtain a residual life prediction value of the motor equipment; repeating the above steps, optimizing the to-be-trained parameters based on an Adam optimization algorithm, and obtaining an optimal residual life prediction model; and inputting an original mon