A Multirate Sensor Information Fusion Strategy for Multitask Fault Diagnosis Based on Convolutional Neural Network

In complicated mechanical systems, fault diagnosis, especially regarding feature extraction from multiple sensors, remains a challenge. Most existing methods for tend to assume that all sensors have uniform sampling rates. However, complex systems use multirate sensors. These upsampling data preprocessing ensure signals at the same scale can cause certain time-frequency features vanish. To address these issues, this paper proposes Multirate Sensor Information Fusion Strategy (MRSIFS) multitask diagnosis. The proposed method is based on multidimensional convolution blocks incorporating multisource information fusion into convolutional neural network (CNN) architecture. Features with different rates raw are run through multichannel parallel framework Additionally, analysis technology used reveal in association between time and frequency domains. simulation platform’s experimental results show model achieves higher diagnosis accuracy than methods. Furthermore, manual selection each task becomes unnecessary MRSIFS, which has potential toward general-purpose framework.